APOPTOSIS IN CHRONIC LYMPHOCYTIC LEUKAEMIA

Transcription

1 PhD Thesis APOPTOSIS IN CHRONIC LYMPHOCYTIC LEUKAEMIA Diane King BSc (Hons) Centre for Mechanisms of Human Toxicity and Department of Pathology University of Leicester

2 UMI Number: U All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. Dissertation Publishing UMI U Published by ProQuest LLC Copyright in the Dissertation held by the Author. Microform Edition ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml

3 This thesis is dedicated to Barrie, Jennifer and Angela King and especially to my husband, James Dodd.

4 ACKNOWLEDGEMENTS I would like to thank my supervisors, Professor G.M. Cohen and Dr. J.H. Pringle for their constructive advice throughout the duration of this study, and Dr. R.M. Hutchinson for his support and guidance. Many thanks also to Mr. R. Snowden, Dr. M. MacFarlane, Mr. D. Brown, Dr. J. Shaw, Mrs. L. Primrose, Mrs. A. Gillies, Mrs. L. Potter and Mrs. T. de Haro for their technical advice and assistance. PUBLICATIONS King D, Cohen GM, Pringle JH, Hutchinson RM: Spontaneous apoptosis in chronic lymphocytic leukaemia as a possible predictor of response to drug therapy. Br. J. Haematol., 1997, 97, Suppl. 1; Abstract 112. King D, Pringle JH, Hutchinson RM, Cohen GM: Processing/Activation of caspases -3, -7 and -8 but not caspase-2 in the induction of apoptosis in B-chronic lymphocytic leukaemia cells. Leukemia, 1998,12; King D, Pringle JH, Hutchinson RM, Cohen GM: Processing/Activation of caspases -3, -7 and -8 but not caspase-2 in the induction of apoptosis in B-chronic lymphocytic leukaemia cells. J. Pathol., 1998,186 SS; Abstract 31.

5 ABSTRACT B cell chronic lymphocytic leukaemia (B-CLL) is the most common adult leukaemia in the western world. The disease is characterised by the accumulation of a CD5+ B cell clone. Drug resistance is a major problem in B-CLL and complete remissions are uncommon. The lymphoaccumulative nature of B-CLL implies that dysregulation of the apoptotic process may be responsible for the development and progression of the disease. B-CLL cells were freshly isolated from patients, and an in vitro apoptosis sensitivity assay was developed using flow cytometric techniques. Initial studies confirmed the existence of spontaneous apoptosis when B-CLL cells were cultured in vitro, and demonstrated a strong correlation between sensitivity to spontaneous apoptosis and sensitivity to apoptosis induced by the chemotherapeutic drug chlorambucil in vitro. Immunoblotting of chlorambucil and prednisolone treated B-CLL cells demonstrated the expression and activation of caspases -3, -7 and -8 in all samples analysed, whilst activation of caspase-2 was seen in cells from only one patient. Activation of caspases -3 and -7 was accompanied by the proteolysis of the DNA repair enzyme, poly (ADP-ribose) polymerase (PARP). Induction of apoptosis and activation of all the caspases was inhibited by the cell permeable caspase inhibitor, benzyloxycarbonyl-val-ala-asp (OMe) fluoromethyl ketone (Z-VAD.fmk). These results demonstrated a key role for the activation and processing of caspases in the execution phase of apoptosis in B-CLL cells. The B cell growth factors interleukin-4 and CD40 were demonstrated to strongly influence survival of B-CLL cells in in vitro culture, and also to modulate the response of the cells to chemotherapeutic drugs. Investigations into Fas induced apoptosis in B-CLL cells demonstrated the expression of the adapter protein, FADD, but no overexpression of the caspase-8 inhibitory protein, c-flip. Additionally, B-CLL cells did not show rapid assembly of the death inducing signalling complex (DISC) in response to stimulation of Fas receptor, implying that these cells may preferentially utilise the Bcl-2- inhibitable Type II (mitochondrial) pathway of apoptosis induction, underlining the important role that Bcl-2 plays in determining the apoptotic sensitivity of B-CLL cells. KEYWORDS Chronic lymphocytic leukaemia; apoptosis; caspases; IL-4; CD40; Fas/CD95

6 LIST OF ABBREVIATIONS ATP = Adenosine triphosphate B-CLL = B cell chronic lymphocytic leukaemia BSA = bovine serum albumin CAGE = conventional agarose gel electrophoresis DISC = death inducing signalling complex EDTA = ethylenediaminetetraacetic acid ELISA = enzyme linked immunoabsorbancy assay FADD = Fas associated death domain protein FIGE = field inversion gel electrophoresis FITC = fluorescein isothiocyanate FLIP = Flice-like inhibitory protein IL-4 = interleukin-4 ISEL = in situ end labelling PARP = poly (ADP-ribose) polymerase PBS = phosphate buffered saline PI = propidium iodide PS = phosphatidylserine RT-PCR = reverse transcriptase PCR SDS = sodium dodecyl sulphate TBS = Tris buffered saline TdT = terminal deoxynucleotide transferase TNF = Tumour necrosis factor TRAIL = Tumour necrosis factor-like apoptosis inducing ligand UP = ultrapure Z-VAD.fmk = benzyloxycarbonyl-val-ala-asp (OMe) fluoromethyl ketone

7 Apoptosis in Chronic Lymphocytic Leukaemia CONTENTS PAGE No 1.0 GENERAL INTRODUCTION Introduction to this thesis Chronic Lymphocytic Leukaemia Diagnosis and Prognosis Clinical staging of CLL Treatment strategies Apoptosis Historical perspectives Morphological and Biochemical characteristics of apoptotic cells The execution phase of apoptosis Mechanisms of induction of apoptosis i Receptor/Ligand signalling pathways ii The mitochondrial pathway of apoptosis induction - role of the bcl-2 family of apoptosis regulators Inhibitors of apoptosis i Inhibitors of receptor/ligand induced apoptosis ii Inhibitor of apoptosis proteins (IAP s) Apoptosis and Chronic Lymphocytic Leukaemia The Bcl-2 family Caspase expression in CLL Growth factor dependency of CLL cells Aims and Objectives MATERIALS AND METHODS Drugs, chemicals and stock solutions Selection of Patients Isolation of B-CLL cells from whole blood Isolation of total lymphocyte fraction Purification of B lymphocytes Determination of apoptosis sensitivity In situ end labelling (ISEL) Annexin V assay for apoptosis One-stage DNA fragmentation analysis gels Field inversion gel electrophoresis (FIGE) Immunoblotting Stimulation of B-CLL cells using anti-cd40 monoclonal antibody and interleukin Analysis of CD95/Fas receptor expression on B-CLL cells Isolation of the death-inducing signalling complex (DISC) 35

8 Apoptosis in Chronic Lymphocytic Leukaemia 3.0 RESULTS 1 - Development of an in vitro apoptosis sensitivity 36 assay for CLL cells 3.1 Introduction Spontaneous apoptosis of CLL cells Apoptotic CLL cells exhibit limited nucleosomal DNA cleavage, but evidence of large DNA fragmentation can be observed The ISEL assay underestimates the percentage of apoptotic CLL cells In vitro sensitivity to spontaneous apoptosis is a predictor of in vitro sensitivity to chemotherapeutic drug-induced apoptosis As patients undergo chlorambucil therapy, the in vivo level of apoptosis can decrease, but the sensitivity of the cells to spontaneous and chlorambucil-induced apoptosis in in vitro culture increases Discussion RESULTS 2 - Processing/activation of caspases -3, -7 and -8, but 53 not caspase-2 in the induction of apoptosis in CLL 4.1 Inhibition of spontaneous apoptosis in CLL cells by Z-VAD.fmk Activation of caspase-3 and caspase-7 in apoptosis of CLL cells Caspase-2 processing does not generally accompany apoptosis of CLL cells Activation of the effector caspases results in the cleavage of PARP Activation of caspase-8 during apoptosis of CLL cells Z-VAD.fmk inhibits the processing of caspases in CLL cells Discussion RESULTS 3 - Studies on survival factors and the Fas signalling 69 pathway in B-CLL 5.1 Introduction Purified B-CLL lymphocytes are more sensitive to apoptosis in vitro culture than unpurified populations of total CLL lymphocytes Culture of B-CLL cells with interleukin-4 and CD40 stimulation results in a reduction in spontaneous apoptosis Culture of B-CLL cells with interleukin-4 or CD40 stimulation increases their resistance to chemotherapeutic drugs B-CLL cells are not sensitive to CD95-induced apoptosis Upregulation of CD95 receptor on B-CLL cells does not increase their sensitivity to apoptosis induced by CD95 ligation B-CLL cells do not over-express the caspase-8 inhibitory protein, c-flip B-CLL cells have all the necessary components to form a death-inducing signalling complex (DISC) upon CD95 ligation, but do not assemble a DISC in response to CD95 stimulation Discussion Survival Factors in B-CLL Investigations into the Fas signalling pathway in B-CLL 95

9 Apoptosis in Chronic Lymphocytic Leukaemia 6.0 GENERAL DISCUSSION 100 REFERENCES 105 APPENDIX 119 LIST OF FIGURES AND TABLES FIGURE 1.1 Mechanism of activation of caspase Pathways leading to cell death Annexin V/PI dot plot to illustrate positioning of quadrant CLL cells were isolated from whole blood samples, labelled using the ISEL technique and analysed flow cytometrically CLL lymphocytes were analysed by agarose gel electrophoretic methods A comparison of the ISEL and Annexin V labelling techniques In vitro sensitivity to spontaneous apoptosis is a predictor of sensitivity to in vitro chlorambucil-induced apoptosis In vivo apoptosis can decrease and sensitivity to spontaneous apoptosis can increase as drug treatment is administered Induction of apoptosis in cells from two patients with CLL assessed by phosphatidylserine extemalisation Induction of apoptosis in a representative B-CLL patient is accompanied by processing of caspase-3 and caspase Activation of caspase-2 only occured in cells from 1 patient Cleavage of PARP accompanies apoptosis in CLL cells Z-VAD.fmk Processing of caspase-8 in CLL cells is inhibited by Z-VAD.fmk Processing of caspase-3 in CLL cells is inhibited by the caspase inhibitor Z-VAD.fmk A Use of CD3+ magnetic beads enriches for CD19+ B cells in cultures of CLL lymphocytes 5.1 B Purified B cells are more sensitive to spontaneous apoptosis than cultures of mixed lymphocytes Interleukin-4 inhibits the induction of spontaneous apoptosis in purified B-CLL cell cultures The effect of stimulation of B-CLL cells with interleukin-4 and antibodies to CD40 is case dependent Stimulation of B-CLL cells with anti-cd40 and interleukin-4 protects against apoptosis induced by chlorambucil B-CLL cells are resitant to apoptosis induced by anti-fas monoclonal antibody. 79

10 Apoptosis in Chronic Lymphocytic Leukaemia 5.7 Fas receptor expression is elevated on B-CLL cells following stimulation with antibodies to CD40 and/or interleukin Upregulation of Fas receptor on B-CLL cells does not increase their sensitivity to apoptosis induced by Fas stimulation. 5.9 Stimulation of B-CLL cells with CD40 does not confer sensitivity to Fas-induced apoptosis B-CLL cells do not overexpress the Fas signalling pathway inhibitory protein c-flip B-CLL cells express the adapter protein, FADD Immunoblotting for FADD on immunoprecipitates of Fas stimulated cells Immunoblotting for caspase-8 on lysates of anti-fas stimulated cells TABLE 2.1 List of patients involved in the study 4.1 Inhibition of spontaneous apoptosis by Z-VAD.fmk 4.2 Clinical information and summary of in vitro apoptosis sensitivity and incidence of caspase activation 25/

11 Chapter 1 Introduction GENERAL INTRODUCTION

12 Chapter 1 Introduction INTRODUCTION TO THIS THESIS Described in this thesis is an investigation into the role played by apoptosis in the etiology and treatment of chronic lymphocytic leukaemia (CLL). This first chapter serves as an introduction to chronic lymphocytic leukaemia, and as a precis of the increasingly complex field of apoptosis research. The final section of this chapter summarises research which has been undertaken to date relating to apoptosis and B- cell chronic lymphocytic leukaemia (B-CLL), and identifies areas where further research might be implicated. Subsequent chapters describe the techniques applied and the results generated during the research project pertaining to this thesis. An appraisal of the main findings, along with a discussion about future directions for research in this field, is included at the end. 1.1 CHRONIC LYMPHOCYTIC LEUKAEMIA Diagnosis and Prognosis B cell chronic lymphocytic leukaemia (B-CLL) is the most commonly occuring adult leukaemia in the West and is a result of accumulation of a lymphocyte clone that is highly resistant to cell death. Since the circulating malignant clone does not undergo apoptosis, the tumour burden increases in volume and in extent of spread. B-CLL can therefore be termed a Tymphoaccumulative disorder (Dameshek, 1967). Increasingly B-CLL patients are diagnosed in an asymptomatic phase. This is most likely attributed to the practice of carrying out blood tests for minor reasons. Diagnosis is made by the presence of lymphocytosis in peripheral blood and bone marrow and will include immunophenotypic evaluation to confirm the presence of characteristic markers. B-CLL lymphocytes express low amounts of surface IgM, or a combination of slgm and slgd (Rozman & Monserrat, 1995). Other surface markers commonly expressed on B-CLL lymphocytes include CD20, and CD23, however, the most common B-CLL marker is CD5 in conjunction with CD 19 (Caligaris-Cappio et at, 1993). 2

13 Chapter 1 Introduction CD5 is a 67 kd glycoprotein which was originally described as a T cell antigen. Only a small subset of normal B cells found at the edge of the germinal centres of human lymph nodes are believed to carry CD5. Additionally, a substantial number of foetal B cells express CD5. In total approximately 5-10% of normal B cells express CD5, so finding a normal counterpart to the B-CLL lymphocyte to use experimentally to examine the lineage of the B-CLL clone has proved to be difficult. Morphologically, the B-CLL lymphocyte is not distinct from normal B cells, if slightly smaller in size. B-CLL cells typically have a high nuclearrcytoplasmic ratio, and nuclear chromatin often appears clumped. Nucleoli are usually indistinct or not visible. Conflicting reports exist regarding the stage of maturation of the malignant cells, but they do appear to be predominantly mature. Any minor variations in shape or size do not appear to correlate with clinical status or progression. The volume of neoplastic B lymphocytes increases with time, so even if a patient is diagnosed in an asymptomatic phase, symptoms will eventually appear. These typically include lymphadenopathy, splenomegaly and hepatomegaly. The white cell count, an important monitor of disease progression, will increase, and anaemia and thrombocytopenia often occur as a result of bone marrow infiltration. Marrow infiltration can be classified into three types, diffuse, where marrow fat and interstitium are replaced by extensive lymphocytosis, focal, characterised by distinct, randomly distributed aggregates of lymphocytes, and interstitial, where the overall marrow architecture is preserved. The median survival is approximately nine years, depending on clinical stage at diagnosis. Good predictors of survival include low clinical stage, low blood lymphocyte counts, and positive bone marrow histopathological findings (low levels of lymphocyte infiltration). Positive response to therapy is also a good prognostic indicator. An indicator of poor prognosis is transformation of B-CLL into large cell lymphoma (Richters syndrome) (Sawitsky & Rai, 1992). Cytogenetic abnormalities are also screened for in B-CLL and found in around 50% of cases (O Brien et al, 1995). An abnormal karyotype can indicate poor prognosis. Karyotypic evolution as the disease progresses is rare. The most commonly found 3

14 Chapter 1 Introduction cytogenetic abnormality is a deletion at 13ql4. Trisomy of chromosome 12, alone or with abnormalities of chromosomes 11 and 14 are also found. No ras mutations (chr 12) have been found to date, and translocation of bcl-2 (chr 14) is relatively rare. However, increased mrna and protein expression of bcl-2 is found in a majority of B-CLL cases, although the mechanism by which this is made possible is unclear, although it has been postulated that hypomethylation of the bcl-2 gene may be involved (Hanada et al, 1993). The retinoblastoma (Rb) gene on the long arm of chromosome 13 is rarely deleted or translocated in B-CLL. When the chromosome 13 breakpoint region was investigated, a high frequency of deletions was discovered at a locus 530 kb away from the Rb gene. These deletions are often homozygous, and could indicate the presence of a new tumour suppressor gene Clinical Staging of B-CLL Two staging systems exist to define disease status, and therefore aid diagnosis, and choice of treatment strategy. That developed by Rai (Rai et al, 1975), distinguishes between five stages of disease :- Rai staging system Stage 0 Stage I Stage II Stage III Stage IV Lymphocytosis in blood and BM. + lymph node involvement. + organ involvement. + anaemia. + thrombocytopenia. The other commonly used system, is that developed by Binet (Binet et al, 1981), which classifies patients into three stages of disease depending on the extent of spread of lymphocytosis (This is the staging system employed in the present study). :- 4

15 Chapter 1 Introduction Binet staging system Stage A Stage B Stage C Lymphocytosis and < 3 areas of lymphoid enlargement Lymphocytosis and > 3 areas of lymphoid involvement Lymphocytosis and < log/dl Hb, or < 100 x 109/ L platelets. Clinical progression is subsequently assessed by monitoring the peripheral blood lymphocyte count, platelet count, lymph node appearance and response to treatment Treatment Strategies The alkylating agent chlorambucil has long been the drug of choice in B-CLL. It is still the most common first-line therapy, and although complete remissions are rare, chlorambucil does reduce the overall tumour burden in the majority of cases. In the asymptomatic phase of the disorder it has been demonstrated that treatment with chemotherapy can actually be detrimental to the patient. For example, patients with early stage B-CLL treated with chlorambucil have a greater risk of developing epithelial cancers, and thus have a reduced survival rate when compared with those patients not receiving early treatment, (Monserrat & Rozman, 1994). As the disease progresses, and symptoms begin to appear, it is usual to treat with chlorambucil, at a dose of mg/kg every 2-3 weeks. Chlorambucil acts by alkylating the N7 position of guanine nucleotides producing adducts in the DNA. Corticosteroids such as prednisolone are often given in conjunction with chlorambucil to counter immune haemolysis. The dose given is in the range mg/m2 /day orally. Prednisolone cytotoxicity is mediated via nuclear receptor interaction. Combination therapies have been the subject of many clinical trials. These include COP (cyclophosphamide, vincristine, prednisone), CHOP (as COP but with doxorubicin), and MOPP (nitrogen mustard, vincristine, procarbazine, prednisone), 5

16 Chapter 1 Introduction amongst others. However, results indicate no great improvement on the response rates for chlorambucil alone (Monserrat & Rozman, 1994). Recently, much research has centered on the new breed of chemotherapeutic drugs, the purine analogues, one of which is fludarabine monophosphate (Astrow, 1996). Fludarabine is transported into the cell by nucleoside transporter proteins. Once inside the cell, fludarabine is phosphorylated by deoxycytidine kinase to F-ara-adenine triphosphate (F-ara-ATP). Because of its resistance to deamination by adenosine deaminase (an enzyme which, incidentally, is naturally depleted in many B-CLL patients (Sawitsky & Rai, 1992)), F-ara-ATP accumulates in the cell. The result of this accumulation is suppression of DNA synthesis by inhibition of DNA polymerase a (O Brien et al, 1995). Other groups have proposed that incorporation of F-ara-ATP into replicating DNA is vital to the activity of this drug (Huang & Plunkett, 1995). The dose given is in the range 25-30mg/m2 for five days, repeated every four weeks for 4-6 cycles. Since fludarabine is a relatively expensive drug to administer clinicians are eager to introduce methods of predicting which patients will respond favourably before treatment commences. Drug resistance is a major problem in B-CLL and may be linked to an increased resistance to apoptosis. Several studies have investigated the expression levels of MDR1 (P-glycoprotein) and glutathione-s-transferases following treatment of B-CLL patients with chlorambucil. One study showed that MDR1 expression was increased following treatment (Perri et al, 1989), and GST expression has been shown to be approximately doubled in a proportion of chlorambucil-treated patients (Schisselbauer et al, 1990). 6

17 Chapter 1 Introduction 1.2 APOPTOSIS Historical Perspectives Apoptosis, or programmed cell death, was first recognised as being distinct from necrosis by Kerr and co-workers in the early 1970 s. Electron microscopy was employed to study some novel histopathological events which had been observed in lysosomes in hepatic ischemia. These events consisted of the formation of discrete, rounded bodies, some of which contained chromatin fragments. Electron microscopy revealed that these bodies contained intact organelles, and that they had arisen by condensation and disruption of an original hepatocyte. This phenomenon, originally termed shrinkage necrosis (Kerr, 1971), was subsequently observed in many different tissues. The name apoptosis was proposed in 1972 (Kerr et al, 1972) and is derived from the Greek word meaning falling off, as in leaves from trees. Apoptosis has since been shown to occur throughout many physiological events, from limb development in the foetus, to cell killing by cytotoxic T lymphocytes Morphological and Biochemical Characteristics of Apoptotic Cells Necrotic cell death is characterised by swelling of the cell and organelles followed by loss of membrane integrity. When this occurs in tissue, inflammation is inevitable. Conversely, apoptosis is characterised by compaction of chromatin into crescent shapes which lie against the nuclear membrane followed by condensation of the cytoplasm which is not accompanied by swelling of the organelles as occurs in necrosis. The cell membrane becomes crenellated or blebbed, and the cell detaches from its neighbours. Eventually, the cell divides up into multiple membrane bound apoptotic bodies, some of which contain chromatin. In tissues, these apoptotic bodies are rapidly cleared, due to the expression of signalling molecules on the cell surface of the apoptotic cell which are recognised by phagocytic cells. As a result of the rapid clearance of dead cells from the tissue, inflammation does not occur. Occuring synchronously with condensation of chromatin is cleavage of doublestranded DNA at nucleosomal intervals resulting in the formation of fragments of 7

18 Chapter 1 Introduction DNA in multiples of approximately 200 base pairs (Bortner et al, 1995). An associated biochemical marker is the production of large fragments of DNA ( kbp), prior to or in the absence of intemucleosomal DNA cleavage. (Oberhammer et al, 1993). A number of techniques for quantifying apoptosis are based upon recognition of DNA cleavage. The in situ end labelling (ISEL) and TUNEL methods can be used on tissue sections or as a flow cytometry method on cell suspensions. Both techniques use the enzyme terminal deoxynucleotide transferase (TdT) to label the cleaved DNA fragments with a tail of digoxygenin-labelled nuclotides. Antidigoxygenin Fab fragments incorporating an colourimetric or fluorescent marker are used to label the cleaved DNA. Other methods of observing apoptotic cells include agarose gel electrophoretic methods. The classic ladder of nucleosomal fragments can be visualised using conventional agarose gel electrophoresis (CAGE), and the formation of the larger DNA fragments can be observed using pulsed field or field inversion gel electrophoresis (FIGE). Since the early 1990 s techniques have been developed using other hallmarks of the apoptotic process in order to quantify cell death, usually flow cytometrically. One of the most widely used techniques is that of Annexin V / propidium iodide labelling (Koopman et al, 1994). One of the molecules on the cell surface of an apoptotic cell, which acts as a signalling marker to phagocytic cells, is phosphatidylserine (PS). PS is a phospholipid usually resident on the inner leaflet of the cell membrane. During apoptosis PS is flipped onto the outer surface of the cell. Use of fluorescein isothiocyanate (FITC)-labelled Annexin V, a protein with high affinity for PS, in conjunction with the red fluorescent dye, propidium iodide (PI) in a dye exclusion role, results in an extremely efficient flow cytometric method for analysing apoptosis and for distinguishing apoptotic from necrotic cells The Execution Phase of Apoptosis Apoptosis can be divided into two phases: an initial condemned phase where cells are committed to die, without any morphological changes, followed by an execution phase when the characteristic biochemical and morphological changes of apoptosis occur, as described above. Apoptosis can occur in response to a variety of stimuli. Chemotherapeutic drugs, irradiation, and triggering of cell surface death receptors 8

19 Chapter 1 Introduction such as Fas (CD95 / Apo-1) and Tumour necrosis factor receptor (TNF-R1) can all result in induction of apoptosis. Whatever the nature of the death signal, or the mechanism by which it is received in the cell, there appears to be a common execution pathway, which results in cell death. Much of our present knowledge regarding this phase of apoptosis comes as a result of studies performed on the nematode worm Caenorhabditis elegans (Yuan et al, 1993). It is known exactly how many cells are produced, and how many die during development of this organism, and the genes which must be mutated in order for the organism to deviate from normal development. The genes ced-3 and ced-4 were demonstrated to be essential for developmental cell death in C. elegans since ced-3 or ced-4 null worms have 131 surplus cells at birth. The mammalian ced 3 homologues are cysteine proteases called caspases (previously ICE-like proteases) for cysteine-aspartate proteases. Caspases are produced as inactive zymogens that require cleavage at the PI position of an aspartate residue in order to attain their active form. They in turn specifically cleave target substrates at the PI of an aspartate residue in a defined amino acid sequence. The caspases are composed of pro-domains followed by two shorter domains, the large and small subunits. Typically activation involves cleavage at the pro-domain/large subunit junction, followed by cleavage at the large/small subunit junction and heterodimerisation of the large and small subunits (Han et al, 1997). A complex consisting of two heterodimers is thought to make up the active enzyme (figure 1.1). Caspases can be divided into two groups, initiator caspases and effector caspases. Initiator caspases preferentially cleave at a (IVL)ExD sequence and typically have long pro-domains. Initiator caspases such as caspase-8 (FLICE/MACH) (Boldin et al, 1996; Scaffidi et al, 1997) and caspase-2 (Ich-1) (Harvey et al, 1997) are involved in apoptosis directed through the membrane signalling molecules Fas (CD95/Apo-l) and TNF-R1. Effector caspases, which include caspase-3 (CPP32) (Nicholson et al, 1995) and caspase-7 (Mch-3) (Fernandes-Alnemri et al, 1995), cleave at a DEVD target sequence and are responsible for cleaving substrates which result in cell death. These targets include poly ADP-ribose polymerase (PARP) (Kaufinann et al, 1993; Casciola-Rosen et al, 1996), nuclear lamins (Rao et al, 1996) and ICAD, the inhibitor of the caspase-dependent DNAse (CAD) which is responsible for cleaving DNA into the characteristic intemucleosomal pattern (Enari et al, 1998). 9

20 Chapter 1 Introduction 32 kd PRO 17 kd 12 kd Caspase-3 zymogen PRO 17 kd 12 kd IStage 1 : Cleavage at 12 kd/17 kd domain junction Stage 2 : Cleavage at pro-domain/17 kd domain junction 17 kd 12 kd 17 kd 12 kd Stage 3 : Dimerisation o f two active caspase-3 enzymes 17 kd 12 kd 12 kd 17 kd Active caspase-3 Figure 1.1 Mechanism of activation of caspase-3 10

21 Chapter 1 Introduction Until recently it was considered that caspase activation was essential for the execution phase of apoptosis. However, certain instances of apoptotic cell death exist where no evidence of caspase activation or intemucleosomal DNA cleavage is seen. Caspaseindependent apoptosis in yeast (Green & Reed, 1998), sperm, and cyclohexamide/staurosporine stimulated chicken erythrocytes (Weil et al, 1998) has been described, although the molecules involved in this type of apoptosis are yet to be described Mechanisms of Induction of Apoptosis 1.2.4i Receptor-Ligand Signalling Pathways Triggering of apoptosis can occur via two mechanisms which are not mutually exclusive. One route for apoptosis initiation is via triggering of the cell surface death receptors Fas (CD95/Apo-l), TNF-R1 or the TRAIL (tumour necrosis-like apoptosis initiating ligand) receptors. Binding of Fas ligand to the Fas receptor initiates trimerisation of the receptor, bring into close contact three intracellular death domains (DD), a region of homology shared with members of the TNF-family, and some of the initiator caspases such as caspase-8 (also known as FLICE ). The receptor death domains are recognised and bound by a death domain containing adapter molecule called FADD (Chinnaiyan et al, 1996). Other adapter molecules which are involved in apoptosis signalling through TNF and TRAIL receptors include RAIDD/CRADD (Duan & Dixit, 1997) and Daxx (Yang et al, 1997). Activation of the receptor/ligand apoptosis initiating pathways triggers a series of events which result in activation of the effector caspases leading to cell death. Activation of caspases via Fas signalling seems to occur by two distinct mechanisms (Scaffidi et al, 1998). Type I is extremely rapid (within seconds) and requires formation of a death inducing signalling complex (DISC). The DISC consists of caspase-8 which is activated in response to Fas signalling by binding to the death effector domain (DED) of the adapter molecule FADD (figure 1.2). Activated caspase-8 then triggers cleavage and activation of caspase-3 and caspase-7, leading to degradation of the cell. Type II Fas signalling does not seem to require the formation of a DISC, but appears to be dependent on mitochondrial activity, in that it can be blocked by Bcl-2 overexpression. Type II 11

22 Chapter 1 Introduction cells appear to activate caspase-3 via the formation of an apoptosome consisting of caspase-9, Apaf-1 and cytochrome c, which is released from mitochondria (figure 1.2). Type II cells show activation of caspase-3 after approximately 60 minutes of Fas stimulation. Activation of the TNF receptor by TNF binding can promote survival or apoptosis depending on the adapter and accessory proteins which bind to the intracellular domain of TNF-R1. TRADD is the primary adapter molecule for TNF signalling and can recruit FADD and caspase-8 to induce apoptosis (Yuan, 1997). Alternatively, the adapter molecule RAIDD/CRADD (Duan & Dixit, 1997) can mediate an NFkB signal in response to binding of the TNF receptor by the kinase RIP (TNF-receptor interacting protein) (Ahmad et al, 1997; Kelliher et al, 1889). Other molecules involved in receptor-mediated apoptosis induction include CARDIAK, which is a RIP-like kinase that can bind and activate caspase-1 via its DED/CARD domain, and is also involved in triggering of Jun kinase and NFkB (Thome et al, 1999). Daxx is a second adapter molecule in the CD95 system which triggers Jun kinase, via activation of apoptosis signal-regulating kinase-1 (ASK). FLASH is another protein involved in receptor-mediated apoptosis. FLASH is a large protein of approximately 220 kb with homology to C. elegans CED-4, which interacts with caspase-8 through a DED domain, and appears to be another Fas DISC component (Imai et al, 1999), although its function is unclear. The Fas and TNF receptor/ligand partemships are just two of a number of related pairings which when triggered can cause induction of apoptosis. TRAIL (TNF-related apoptosis initiating ligand) can bind to a number of receptors. Some of these, such as DR4 and DR5, when bound by TRAIL, lead to induction of apoptosis. Others, termed decoy receptors such as DcRl and DcR2, lack the intracellular death domain necessary for apoptosis induction. Binding of TRAIL to these receptors, therefore, does not induce apoptosis (French & Tschopp, 1999 for review). TRAIL has been shown to induce apoptosis in cells from a range of haematological neoplasms, including chronic lymphocytic leukaemia (Snell et al, 1997). 12

23 CELL MEMBRANE Fas Ligand Fas receptor A FADD Chemotherapeutic drugs, growth factor withdrawal, mitogenic signals Ceram ide CYTOPLASM c-flip p- Activated Caspase-8 Activated Caspase-3 Cytochrome c Activated caspase-9, datp, Apaf-1 apoptosome MITOCHONDRION NUCLEUS Morphological nuclear events Figure 1.2 Pathways leading to cell death tcell Death i

24 Chapter 1 Introduction 1.2.4ii The Mitochondrial Pathway o f Apoptosis Induction - Role o f the Bcl-2 Family o f Apoptosis Regulators The mitochondrial pathway of apoptosis induction can be triggered via Fas signalling as described above, or directly via triggering loss of mitochondrial membrane potential (Susin et al, 1997). Cytochrome c, released from the mitochondria due to the opening of a permeability transition pore, triggers the mammalian homologue of the C. elegans ced-4 gene product, Apaf-1, in the presence of datp, to activate caspase-9 in a complex which is termed the apoptosome (Li et al, 1997). Activated caspase-9 subsequently activates the effector caspases such as -3 and -7. Whether or not the loss of mitochondrial membrane potential is linked to the release of cytochrome c is unclear at the present time. One mediator of the mitochondrial apoptotic pathway, ceramide, is produced by a number of apoptotic initiating agents and accumulates at the mitochondria, where it can directly bind and release cytochrome c (Garcia-Ruiz et a l 1997). Members of the Bcl-2 family exert some of their effects at the mitochondia. Bcl-2 is a 26 kd protein located at 18q21, originally identified during analysis of the t(14; 18) breakpoint in follicular lymphoma (Tsujimoto et al, 1985). It is homologous to the apoptosis inhibitory ced-9 gene product in C. elegans. A family of related proteins has since been discovered, some with anti-apoptotic effects like Bcl-2, and others, such as Bax, with pro-apoptotic activity (Oltvai et al, 1993), all containing regions of homology termed the BH domains. Since induction of apoptosis by a wide variety of stimuli, including anti-cancer drugs, UV irradiation and receptor-ligand signalling pathways can be blocked by overexpression of Bcl-2 (Cory, 1995), it appears that Bcl-2 must act at a point in the apoptotic process that is common to many pathways. Investigations into the mechanism of action of Bcl-2 and related proteins has centered on their activity in mitochondria during triggering of apoptosis. Early studies on Bcl-2 localised the protein to the mitochondrial outer membrane, particularly at points where the outer and inner membranes formed a pore. The structure of Bcl-2 and a related protein, anti-apoptotic B c1-xl (Boise et al, 1993) resembles that of certain bacterial toxins which can insert into the mitochondrial membrane and may influence ion channel activity (Kroemer, 1997). Bcl-2 and Bcl-XL overexpression inhibits the 14

25 Chapter 1 Introduction change in mitochondrial membrane potential induced by chemotherapeutic agents (Decaudin et al, 1997) thus contributing to chemoresistance, and more recent studies have demonstrated that anti-apoptotic Bcl-2 interacts with pro-apoptotic Bax on the mitochondrial membrane to prevent activation of downstream caspases, and may remove Bax to the endoplasmic reticulum to prevent Bax-induced release of cytochrome c (Bomer et al, 1999). The important role that Bax plays at the mitochochondria is underlined by the finding that overexpression of Bax can induce mitochondrial permeability transition (Pastorino et al, 1998). The relevance of permeability transition to the release of cytochrome c is a matter under much debate, the two events may be closely related, but reports of cytochrome c release prior to loss of mitochondrial membrane potential have been made (Reed et al, 1998). Members of the Bcl-2 family are often involved in amplification or propagation of the apoptotic signal. In 1998, two pathways for Fas induced apoptosis were reported (Scaffidi et al, 1998) (as described in section 1.2.4i of this chapter). In Type I signalling, characterised by the rapid formation of a DISC leading to activation of significant amounts of caspase-8 and subsequently caspase-3, Bcl-2 is poorly effective as an inhibitor. In Type II signalling, characterised by activation of small amounts of caspase-8, Bcl-2 can act as a potent inhibitory factor, indicating that mitochondria may be involved in this type of apoptosis induction. The link between activation of caspase-8 and mitochondrial events has been shown to be due to the activation of the Bcl-2 family member, BID. BID is a 24 kd, BH3 domain containing protein which has cleavage sites for caspase-8 and granzyme B. Cleavage of BID at Asp 59 by caspase-8 removes an amino terminus inhibitory domain, to yield a potently pro-apoptotic protein (Li et al, 1998, Luo et al, 1998). Truncated BID (tbid) was shown to translocate from the cytoplasm to the mitochondrial and nuclear membranes where it could induce clustering of mitochondria, loss of mitochondrial membrane potential and release of cytochrome c, and eventually cell shrinkage and nuclear condensation (Li et al, 1998). In this way, the Fas induced apoptotic signal can be amplified by release of mitochondrial factors. A second way in which an apoptotic signal can be amplified was revealed by Kirsch and co-workers (Kirsch et al, 1999). Earlier reports had demonstrated that when recombinant anti-apoptotic Bcl- 2 was cleaved at Asp 34 by caspase-3, a 23 kd pro-apoptotic fragment of Bcl-2 was produced (Cheng et al, 1997). Later studies by the same group identified caspase-3 as 15

26 Chapter 1 Introduction the factor required for Bcl-2 cleavage in Fas treated Jurkat cells, and demonstrated the co-localisation of intact and cleaved Bcl-2 to the mitochondria. There they demonstrated that cleaved Bcl-2 could promote release of cytochrome c from the mitochondria in a manner similar to Bax (Kirsch et al, 1999). The authors postulate that cleavage of Bcl-2 at Asp 34, removing the amino terminal BH4 domain, may expose the pro-apoptotic BH3 domain, similar to the effect of caspase-8 cleavage of BID. Thus cleavage of Bcl-2 may act as part of a positive feedback loop, amplifying the apoptotic signal at a mitochondrial level, by promoting release of cytochrome c and subsequent activation of caspase-9 and caspase-3. The release of cytochrome c from mitochondria following an apoptotic stimulus has been well established. However, it has since been reported that a number of other factors are released into the cytosol from mitochondria following induction of apoptosis. These factors include caspase-2 and caspase-9, and a 50 kd apoptosis inducing factor (AIF) (Susin et al, 1999). Bcl-2 may therefore also be involved in preventing the release of caspase-9 and cytochrome c from mitochondria in the absence of an apoptotic stimulus, thereby inhibiting apoptosome formation and activation of downstream caspases such as caspase Inhibitors of Apoptosis Since cancer can be viewed as a situation in which the tumour cells are resistant to cell death, molecules which inhibit apoptosis may play an important role in development of malignancies. Accordingly, much work has been done to identify and characterise the expression of molecules such as these i Inhibitors o f Receptor/Ligand -induced Apoptosis The identity of several proteins which may act at the proximal stage of apoptosis induction have been identified. The first of these was c-flip (Irmler et al, 1997; Rasper et al, 1998) (also called Casper, I-FLICE, FLAME-1, CLARP, MRIT and Usurpin-a), a FLICE-like inhibitory protease which can block apoptosis signalling 16

27 Chapter 1 Introduction through the CD95/Fas system. c-flip has two death effector or caspase-recruitment domains (DED/CARD s) at its amino terminus, which enable binding of c-flip with caspase -8 and FADD in the Fas-induced DISC. Interaction of c-flip with activated caspase-8 causes cleavage and activation of c-flip and results in the c-flip/caspase- 8 interaction becoming more inhibitive. c-flip was found to be predominantly expressed in lymphoid and muscle tissues, and high levels of the protein were discovered in melanoma cells (Irmler et al, 1997). The extent of expression of c-flip in B-CLL cells has not previously been reported. The silencer of death domain proteins (SODD s) are a second class of inhibitor proteins (Jiang et al, 1999). When complexed with the death domain of TNF-R1, SODD keeps the receptor in a monomeric form. Ligation of TNF-R1 by TNF, causes SODD to dissociate, allowing formation of the TNF death inducing signalling complex consisting of the adapter molecules TRADD and TRAF-2 (Yuan, 1997), and the kinase RIP (Kelliher et al, 1998). A situation could therefore exist where a dominant negative mutation of a SODD could leave the receptor trapped in a nonfunctional state. The existence of SODD s in the TNF receptor-ligand system may indicate that similar proteins exist in the Fas system, although none have been described to date ii Inhibitor o f Apoptosis Proteins (IAPfs) IAP s are a family of inhibitor of apoptosis proteins (Liston et al, 1996) originally identified in baculovirus systems, but with mammalian homologues. Mammalian IAP s typically contain one or more baculovirus IAP repeat sequence, and have a carboxy terminus RING finger domain. The IAP s, which in humans include c-iap 1, c-iap 2, x-iap and n-iap, bind to and block the activity of TRAF-2, a protein involved in TNF receptor-mediated activation of NF-xp. Survivin, another IAP (Ambrosini et al, 1997) differs from the original IAP s in that it contains only one baculovirus IAP repeat, and has no carboxy terminal RING finger. Survivin is expressed during foetal development, but is not present in terminally differentiated adult tissue. However, abundant survivin has been detected in several 17

28 Chapter 1 Introduction common human cancers, including breast, colon, prostate and lung, in vivo (Kawasaki et al, 1998). Survivin was also expressed in approximately 50% of high grade non- Hodgkin s lymphomas, but not in low grade cases, indicating that this inhibitor may be important in regulating apoptosis sensitivity in lymphoid malignancies. 1.3 APOPTOSIS AND CHRONIC LYMPHOCYTIC LEUKAEMIA The Bcl-2 Family Due to the discovery that Bcl-2 is overexpressed in a majority of B-CLL cases, many groups have performed investigations in this area. Methods of assessing Bcl-2 expression in B-CLL by different groups range from western blotting of protein samples, to RT-PCR and flow cytometric techniques. It is frequently reported that Bcl-2 gene translocations are extremely rare events in B-CLL, and yet mrna and protein levels of this factor are elevated in the majority of cases. No satisfactory explanation exists, as yet, to explain this phenomenon. Since Bcl-2 is overexpressed in the B-CLL lymphocyte, this may be a possible reason for accumulation of the malignant B-CLL clone. Several groups have investigated this possibility. Gottardi et al (1995) confirmed overexpression of Bcl-2 in seven B-CLL cases by northern blotting and RT-PCR, and postulated that defective apoptosis due to this overexpression of Bcl-2 may be one reason why B-CLL lymphocytes appear as a quiescent cell population, accumulating in the Go phase of the cell cycle. In a later report, the same group analysed twenty-three CLL cases for expression levels of Bcl- 2, Bax, and the two Bcl-X splice variants (Gottardi et al, 1996). They again found high level expression of Bcl-2 in the majority of cases, and in addition discovered elevated Bax protein expression in a proportion of cases. Regarding the expression of B ci-x l and Bcl-Xs, the pattern appears more sketchy, with Bcl-XL more frequently expressed at a higher level than Bcl-Xs. In all, they found that levels of the apoptosis inhibitors, Bcl-2 and Bcl-XL, were raised in a significantly higher proportion of cases than were the apoptosis potentiators, Bax and Bcl-Xs, thus pushing the cells towards an apoptotic block. 18

29 Chapter 1 Introduction A second study examined the expression of Bcl-2 and Bax in relation to in vitro survival of B-CLL lymphocytes, and clinical progression of the patient. They concluded that the ratio of Bcl-2 to Bax was an important factor in determining apoptotic sensitivity in B-CLL (Aguilar-Santelises et al, 1996). This finding was mirrored by that of Thomas and co-workers (1996) who also looked at the importance of this ratio in determining apoptotic sensitivity to chemotherapeutic drugs. Of a panel of eighteen patients, 88% of those with an intermediate to high ratio of Bcl-2 to Bax were in vitro drug resistant, indicating the importance of these factors in B-CLL. More recent work has indicated that drug resistance in B-CLL may be due to selection of subclones which express high levels of Bcl-2 relative to their Bax content (Pepper eta l, 1999) Caspase expression in B-CLL The cellular and tissue distribution of different caspases and their substrate specificity is generally not known, although this is clearly important as exemplified by the inability of some neuronal but not other cells to undergo apoptosis in caspase-3 7' mice (Kuida et al, 1996). Multiple species of caspases-3 and -6 appear to be the major pool of activated caspases in various tumor cell lines induced to undergo apoptosis (Faliero et al, 1997). In addition to caspases-3 and -6, caspases-2 and -7 are also activated in human monocytic tumor cells induced to undergo apoptosis by various stimuli (MacFarlane et al, 1997). Recently it has been shown that peripheral blood mononuclear cells from B-CLL patients are caspase-3 immunopositive (Krajewski et al, 1997) and that glucocorticoid-induced apoptosis of B-CLL lymphocytes requires protease activation and is accompanied by cleavage of PARP and lamin Bi together with the loss of caspase-3 (Bellosillo et al, 1997; Chandra et al, 1997). The importance of caspase activation in spontaneous and chemotherapy-induced apoptosis in leukaemic cells from patients with B-CLL has not previously been reported. 19

30 Chapter 1 Introduction Growth Factor Dependency of B-CLL Cells The characteristic spontaneous apoptosis which occurs when B-CLL cells are cultured in vitro has been reported previously (Collins et al, 1989). Since removal of B-CLL cells from their normal environment appears to trigger the cells into apoptosis, it would seem that the cells are stimulated in vivo by one or more survival factors, which are not provided by in vitro culture in standard media. CD40 is a 45-50kD glycoprotein which is expressed on the surface of B cells at all stages of development, with the exception of terminally differentiated plasma cells. CD40 is a member of the TNF/NGF family of receptors, which also includes Fas/CD95. CD40 ligand (CD40L) is a 35kD glycoprotein, expressed on activated T lymphocytes, follicular and dendritic cells within the germinal centres. Ligation of CD40 receptor by CD40L requires receptor trimerisation, and activates a pathway that results in isotype switching and clonal expansion during the B cell maturation process. CD40 ligation has been demonstrated to rescue isolated germinal centre B cells from apoptosis and has been reported to rescue B-CLL cells from spontaneous apoptosis in in vitro culture. One way in which CD40 might increase protection from apoptosis is by upregulating B ci-x l expression (Ghia et al, 1996). Planken and collegues (Planken et al, 1996), used the CD40 system to induce proliferation of B- CLL cells in vitro. Using CD40-expressing CD32L cells as a feeder layer, and IL4 as a media supplement, they demonstrated proliferation of cells from B-CLL patients. CD40 activation of B-CLL cells has also been used as a method of inducing upregulation of Fas receptor (Wang et al, 1997). Interleukin-4 (B cell growth factor) sources include T cells, monocytes and bone marrow stroma. It is known to stimulate B cell growth and differentiation and antigen presentation. It has been demonstrated that B-CLL cells express either high or low affinity receptors for IL4 at levels comparable to normal B cells (Gileece et al, 1993) Addition of IL4 to cultures of B-CLL cells has been shown to reduce the level of spontaneous apoptosis and to protect against hydrocortisone-induced apoptosis, possibly by upregulating Bcl-2 expression (Danescu et al, 1992). Other techniques for stimulating B-CLL cell growth in in vitro culture include interleukin-2 in combination with mitogens such as Staphylococcus aureus Cowan I (DeFrance et al, 1991), or 20

31 Chapter 1 Introduction pokeweed mitogen (Mainou-Fowler et al, 1995). In both of these studies, IL2 was shown to have a proliferative effect, whilst IL4 had an anti-apoptotic effect. Other growth factors may be important in B-CLL, some of which will be discussed later in this thesis. 1.4 Aims and Objectives The overall purpose of this study was to examine the significance of apoptosis in B cell chronic lymphocytic leukaemia. The fact that B-CLL cells accumulate due an inability to undergo cell death (Dameshek, 1967) points to the importance of understanding the nature of the apoptotic block in B-CLL cells. In addition, since drug resistance is a major problem in B-CLL, this may also be related to the inability of the cells to undergo apoptosis. Research into apoptosis, and the causes and mechanisms of resistance to apoptosis, encompasses a wide range of controlling factors to be investigated. Regarding B-CLL, much research had been done on the role of the Bcl-2 family of apoptosis regulating proteins. This work had been done in response to the finding that B-CLL cells express increased amounts of anti-apoptotic bcl-2. Work in this area has linked the ratio of Bcl-2:Bax to apoptosis sensitivity (Thomas et al, 1996; Aguilar-Santelises et al, 1996, Pepper et al, 1996), but only partially explains the resistance of B-CLL cells to apoptosis induction by chemotherapeutic drugs. In order to contribute to research in this field, it was decided to investigate apoptosis in B-CLL from a variety of different directions. The initial aim of this study was to successfully apply current methods for analysing apoptosis in cell lines to clinical samples with a view to developing a method by which cells from B-CLL patients could be analysed for their sensitivity to apoptosis induced by chemotherapeutic drugs. To then define what factors could influence the apoptosis sensitivity of B-CLL cells was the aim of further research in this study. The identification of caspases as the machinery proteins involved in the execution phase of apoptosis led to a series of experiments being performed to examine the nature of caspase expression and activation in B-CLL cells. The objective of this research was to discover whether B- CLL cells possessed the caspases necessary for apoptosis, and whether these caspases 21

32 Chapter 1 Introduction could be activated in response to stimulation of B-CLL cells with chemotherapeutic drugs. In response to findings produced by this work, the study progressed onto investigating the Fas receptor/ligand signalling pathway of apoptosis induction in B- CLL cells. The aim of this research was to analyse the nature of Fas resistance in B- CLL, by investigating the expression and activation of the various caspases, adapter molecules and inhibitor proteins involved in Fas-induced apoptosis. In conjunction with this work, analysis of the dependency of B-CLL cells on stimulation from survival factors, and the effect that the such stimulation might have on resistance to chemotherapy was investigated. In summary, the main objectives of this study were :- To develop an assay for analysing apoptosis in freshly isolated B-CLL cells. To determine the significance of apoptosis sensitivity in B-CLL and to analyse apoptosis following treatment of B-CLL cells with chemotherapeutic drugs. To investigate the nature of apoptotic resistance in B-CLL, by analysing expression of proteins involved in the apoptotic pathway. To determine whether survival factor dependency plays a significant role in determining the sensitivity of B-CLL cells to apoptosis. 22

33 Chapter 2 Materials and Methods MATF.RIAT.S AND METHODS 23

34 Chapter 2 Materials and Methods 2.0 MATERIALS AND METHODS 2.1 Drugs, chemicals and stock solutions Laboratory stock solutions of phosphate buffered saline (PBS) (0.15 M NaCl (Fisher Scientific, Loughborough, UK), 0.01 M Na2HP0 4 (Fisher Scientific), 2.5 mm NaH2P0 4 2H2O (Hopkins & Williams, Essex, UK), and Tris-buffered saline (TBS) (0.05 M Tris (GibcoBRL, Paisley, Scotland), 0.15 M NaCl (Fisher Scientific)) were used as indicated. Unless stated otherwise all drugs and chemicals were obtained from Sigma Chemical Co. (Poole, Dorset, UK). Prednisolone-21-hemisuccinate was stored as a 0.1 M stock solution in PBS, chlorambucil as a 3 mm solution in ethanol, staurosporine as a 0.2 mm solution in dimethylsulphoxide (DMSO) and fludarabine as a 35 mm solution in dimethyl formamide (DMF). All drugs were stored at 20 C in small aliquots to avoid repeated freezing and thawing. The recommended maximum storage periods at this temperature were not exceeded. Dilutions prior to use were made in 1 x PBS, so that the final concentration of diluent never exceeded 0.1 %. 2.2 Selection of Patients Local ethical committee approval was obtained prior to this study commencing. Patients were assessed according to the Binet staging system, and were sourced from Haematology outpatients clinics. Patients who were undergoing a course of treatment and those who had completed a course within the previous 4 weeks were excluded from the study. Table 1 lists the patients from whom blood samples were obtained for analysis during the study. Of the 44 patients analysed, 55% were male and 45% were female. The ages of the patients ranged from 40 to 94 years, the average age being 70.1 years. The majority (63%) of patients had received no previous treatment for their condition at the time of sample collection. Of the patients who had been treated with chemotherapeutic drugs nine had received chlorambucil alone, four had received 24

35 Chapter 2 Materials and Methods chlorambucil and prednisolone in combination, two had received fludarabine alone, and one had received fludarabine in combination with prednisolone. Patient no. Sex/Age Binet stage WBC count Previous (x 109/L) treatment 11 M/94 A 15.3 none 21 M/85 A 42.0 none 31 M/77 C 3.7 FLD 41 F/73 A 38.2 CHL 51 M/72 A 8.5 PD/FLD 61 F/66 A 51.8 none 71 M/76 C 8.1 PD/CHL 81 F/60 A 68 none 91 F/62 B 35.8 CHL 101 F/84 ND ND none 111 M/67 A 41.7 none 121 M/44 A 15.3 CHL 1 A F/73 A 38.2 CHL 2 A F/80 A 17.1 none 3 A M/74 A 23.1 none 4 A M/72 A 49.7 CHL 5 A M/75 A 20.7 none 6 A F/54 A 29.0 none 7 A M/48 A 38.9 none 8 A M/72 A 30.2 none 9 A M/52 A 9.1 FLD 10 A M/91 C 24.5 none 11 A F/74 B 14.2 CHL 12 A F/67 B 17.5 none 13 A M/62 A 34.3 none 25

36 Chapter 2 Materials and Methods Patient no. Sex/Age Binet stage WBC count Previous (x 109/L) treatment 14 A F/60 A 8.0 none 15 A M/51 A 24.6 none 16 A M/81 A 20.5 CHL 17 A M/71 A 12.7 none 18 A F/84 B 22.7 PD/CHL 19 A M/77 C 7.2 PD/CHL 20 A F/85 A 47.7 none 21 A M/50 B 23.7 none 22 A M/81 C 11.3 PD/CHL 23 A F/75 A ND CHL 24 A F/72 A 10.9 none 25 A F/82 A 66.1 none 26 A M/59 A 22.2 none 27 A F/63 A 29.9 none 28 A F/74 A 39.1 CHL 29 A F/87 A 42.4 none 30 A F/40 A 9.6 none 31 A M/79 A 35.9 none 32 A M/60 A 12.9 none Table 2.1. List of patients involved in the study (PD = prednisolone, CHL = chlorambucil, FLD = fludarabine). White blood cell counts (WBC) and Binet stage refer to patients status at time of sampling. ND indicates not determined. 26

37 Chapter 2 Materials and Methods 2.3 Purification of B-CLL cells from whole blood Isolation of whole lymphocyte fraction Whole blood was collected into K+/ EDTA or Heparin tubes (Sarstedt, Leicester, UK), and layered over Ficoll (Pharmacia Biotech, Sweden). Following centrifugation for 20 minutes at 400g, the lymphocyte layer was aspirated off, and cells were washed twice in culture media (as described below), in order to remove any remaining Ficoll. Freshly isolated lymphocytes were suspended in culture medium (RPMI 1640 (GibcoBRL, Paisley, Scotland), 10% foetal calf serum (FCS) (Sigma), 1% penicillin/streptomycin (GibcoBRL)) and incubated at 37 C for 20 minutes in plastic culture flasks in order to remove adherent cells (macrophages, monocytes). Cells were resuspended at an average density of 1 x 106 cells /ml. (Cells were counted using a haemocytometer, and viability was assessed by Trypan Blue exclusion) Purification of B lymphocytes Lymphocytes isolated as described above were washed in 1 x PBS/2% FCS, and resuspended in PBS / FCS at an average density of 10 x 106 cells/ml. Dynabeads M- 450 CD3 (Dynal, Oslo, Norway) were added to achieve a final ratio of dynabeads: 7 target T cells of at least 4:1, and a minimum concentration of dynabeads of 2 x 10 / ml. The cells and dynabeads were incubated at 4 C for 30 minutes with rolling action. The magnetic beads (and bound CD3+ T cells) were removed by placing the tube of cells into a magnetic particle concentrator (Dynal MPC-1). Unbound cells (CD3* B cells) were pipetted off, washed and resuspended in media (as previously described) at an average density of 1 x 106 cells/ml. Purity of the resulting cell population was determined flow cytometrically using antibodies against the B cell marker CD 19 (CD19-FITC, Dako, High Wycombe, Bucks), and the T cell marker CD3 (CD3-RPE, Dako) at 5pi per 106 cells. Flow cytometry is the measurement of cells in a flow system which has been designed to deliver cells in single file past a laser beam. Fluorescence and scattered light are recorded for each cell and data is depicted 27

38 Chapter 2 Materials and Methods graphically in the form of dot-plots and histograms. From these plots, populations of cells with similar characteristics can be identified and quantified (see figure 5.1, page 71, for an example dot-plot resulting from use of the antibodies described above). 2.4 Determination of Apoptosis Sensitivity Freshly isolated cells were incubated in six-well plates at 37 C in an atmosphere containing 5% CO2. Samples of the cells were incubated in the presence or absence of chlorambucil (3-15 pm), prednisolone ( pm), fludarabine (3 pm), staurosporine (0.2 pm), or anti-fas monoclonal antibody (CH-11) (0.5 pg/ml) (Immunotech, Marseilles, France) (Yonehara et al, 1989) as indicated in the text. For analysis of caspase activation, duplicate samples of untreated and drug-treated cells were pre-incubated for 30 min with 100 pm Benzyloxycarbonyl-Val-Ala-Asp (OMe) fluoromethyl ketone (Z-VAD.fmk), a cell permeable caspase inhibitor (Enzyme Systems Products, Dublin, CA). At specified times samples were analysed for apoptosis, initially using in situ end labelling (ISEL) and agarose gel techniques for visualising DNA fragmentation, and later by Annexin V labelling. 1 x 106 cells from each culture were washed in ice-cold PBS in order to remove media and serum, and were stored at 80 C as cell pellets for immunoblotting. 2.5 In situ end labelling (ISEL) General Principles ISEL incorporates digoxygenin-labelled nucleotides to the ends of DNA strand breaks produced by endonuclease cleavage of DNA during the final stages of apoptosis. The enzyme utilised is terminal deoxynucleotide transferase (TdT), which adds nucleotides to the 3' hydroxyl group of a DNA strand, without the need for a template. The strand breaks are visualised by the addition of an anti-digoxygenin-fab fragments with a fluorescein isothiocyanate (FITC) label. The cells are also labelled with propidium iodide (PI), a red fluorescent DNA staining dye in order to quantify DNA content. Flow cytometry is used to analyse the resulting cell populations. 28

39 Chapter 2 Materials and Methods Method 1 x 106 cells were fixed in 1ml 1% formaldehyde/pbs for 15 minutes on ice, washed in two changes of ice-cold PBS, and resuspended in 1ml 70% ethanol/pbs for storage at 4 C until required. For the labelling reaction, the cells were washed first in 1 x TBS, resuspended in 100 pi of reaction mixture (80 pi UP water, 20 pi 5x TdT buffer, 2 pm digoxygenin-ll-dutp (Boehringer Mannheim, Germany), and 10 U TdT (GibcoBRL, Paisley, Scotland)), and incubated for 35 mins at 37 C. To stop the labelling reaction the samples were removed immediately to ice for 3 mins. The cells were then washed in staining buffer (1% of 10% Triton X-100, 20% 20x standard saline citrate (SSC), plus 5% w/v bovine serum albumin (BSA)) and resuspended in 250 pi labelling mixture (250 pi staining buffer and 0.1 pg anti-digoxygenin-fitc antibody (Boehringer Mannheim)) in which they were incubated for a further 35 mins at 37 C. Following centrifugation at 400g for 5 minutes the samples were resuspended in 1ml propidium iodide (PI) (50 pg/ml), and incubated overnight at 4 C. Flow cytometric analysis was performed the following day on a Becton-Dickinson FACScan, using Lysys II software. A dot-plot of FL2-Area (DNA content) vs FL1- Height (FITC-labelled apoptotic cells) was produced for each sample. Regions were set on FITC-positive and FITC-negative cells for the freshly isolated cell sample in each case, and remained in place for all further analysis on that patient (see figure 3.1, page 38, for an example dot-plot which demonstrates the positioning of the quadrant). Values for the percentage of FITC-positive apoptotic cells in the sample were generated from these regions. Typically 5000 events were recorded per sample. 2.6 Annexin V assay for apoptosis General Principles The Annexin V assay measures a cell surface alteration which occurs during apoptosis. Phosphatidylserine (PS), a membrane phospholipid, is translocated from the inner leaflet to the outer leaflet of the cell membrane. The externalised PS assists recognition of the apoptotic cell by phagocytic cells to aid rapid clearance by engulfment. Annexin V is a Ca2+ -dependent phospholipid binding protein with high affinity for PS. When conjugated with a FITC label, this protein can be used as a 29

40 Chapter 2 Materials and Methods probe for externalised PS by flow cytometry. The cells can be dual labelled with propidium iodide, which acts as a dye exclusion assay for necrotic and secondary necrotic cells. This combination enables a distinction to be made between apoptotic and secondary necrotic cells. An ideal result using this technique gives three populations of cells. Viable cells have minimal staining with both fluorochromes (Pr / FITC ), apoptotic cells are positive for Annexin V-FITC (PI7 FITC+) and secondary necrotic cells (late stage apoptotic cells) stain highly with both fluorochromes (PI+/ FITC+). Method 1 x 106 cells were micro-centrifuged at 400g and then re-suspended in 1ml Annexin V buffer (10 mm HEPES/NaOH ph 7.4, 150 mm NaCl, 5 mm KC1, 1 mm MgCl2, 1.8 mm CaCh). Annexin V-FITC (Bender MedSystems, Vienna, Austria) was added to a final concentration of 1.5 pg/ml. After incubation at room temperature for 8 min, propidium iodide was added (50 pi of 50 pg/ml stock in PBS), and the cells incubated for a further 2 min at room temperature before being placed on ice prior to flow cytometric analysis on a Becton-Dickinson FACStar Plus, using Lysys II software. A quadrant was set on each FL1-H vs FL2-H (Annexin V vs PI) dot plot as illustrated in figure 2.1. Apoptotic cells were identified as Annexin V+ / PI" cells. Secondary necrotic (Annexin V+/ PI*) cells were included in the dead cell count as these were identified as late stage apoptotic cells. Typically 5000 events were recorded per sample. 30

41 Chapter 2 Materials and Methods ANNEXIN V-FITC (3) v s PI (4) <D Secondary necrotic cells A nnexin V+/PI+ 3 ohh g.2 3 '3. 2 Oh <N -J Oh & r: Viable cells... j \.. =: : A poptotic cells l ^ ij l^ ^ '.- A r m e x m V+/PI- 10 FL1-H (Annexin V-FITC) Figure 2.1 Annexin V FITC/Propidium iodide dot plot to illustrate positioning of the quadrant used to quantify cell death in the cell population. 2.7 One-stage DNA fragment analysis gels. Essentially as described by Sorensen et al (1988), this technique enables visualisation of oligonucleosomal DNA cleavage without the need to extract DNA using organic solvents. Briefly, 1 x 106 cells were resuspended in 15 pi of sterile ultrapure (UP) water. 9 pi loading buffer (0.02% bromophenol blue, 40% glycerol in 1 x TBE) and 6 pi RNase A (50 mg/ml) was added and the samples were incubated at room temperature for 30 minutes. A 1.8% agarose gel (UltraPure Agarose, GibcoBRL) (in 1 x TBE) was prepared, and a slice cut from the area above the sample wells. A digesting gef of 0.8% agarose (in 1 x TBE, 2 % SDS), supplemented with proteinase K (50 pl/ml) was poured into the gap above the wells. Once the samples were loaded, the gel was run in 1 x TBE, and bands were visualised by staining for 30 minutes in ethidium bromide (50 pg/100 ml) followed by a 60 minutes de-staining in UP water. 31

42 Chapter 2 Materials and Methods 2.8 Field inversion gel electrophoresis. Essentially as described by Brown et al (1993) and Anand & Southern (1990). Field inversion gel electrophoresis (FIGE) is a form of pulsed field gel electrophoresis. FIGE allows large fragments of DNA up to 700 kb to be resolved on an agarose gel, by alternating the orientation of the electrical field by 180, whilst ensuring forward mobility by increasing the length of the forward pulse over the reverse pulse. In order to protect the DNA from shear stresses, and to ensure support of the DNA fragments in an agarose matrix at all times, the cells were embedded in low melting point agarose plugs (Agarose L, Pharmacia). The plugs were incubated in NDS (1% Lauryl sarcosine, 0.5 M EDTA, 10 mm Tris) and Pronase (1 mg/ml) (Sigma) for 48 hours at 50 C, rinsed in NDS followed by three changes of Tris-HCl ph 8.0, before being loaded on a vertical 1.5 mm thick 1% agarose gel (NA Agarose, Pharmacia). The gel was run overnight using a pulse controller (PC 750, Hoeffer Scientific Instruments), before being stained with ethidium bromide and viewed under UV light. 2.9 Immunoblotting Frozen cell samples were lysed in a 10% SDS sample loading buffer containing bromophenol blue. Electrophoresis was performed on either a 20 cm x 20 cm Flowgen apparatus (Flowgen, Ashby de la Zouch, Leics, UK) using 7 % or 13 % resolving gels with a 4 % stacking gel (Sambrook et al, 1989), or on a 10 cm x 10 cm Novex Mini- Cell II apparatus (Novex, Germany), using 4-12 % pre-cast gradient gels. Samples were run on SDS-polyacrylamide gels as described before being blotted onto nitrocellulose membranes (Hybond C-extra, Amersham, Bucks). Membranes were blocked for 60 min in 5% non-fat dried milk in Tris buffered saline containing 0.1% Tween 20. The membranes were incubated with the primary antibody for 1 h at room temperature followed by washing with Tris buffered saline containing 1% Tween-20 and then incubated with horseradish-peroxidase conjugated secondary antibody (rabbit anti-mouse IgG or goat anti-rabbit IgG, Dako, High Wycombe, Bucks) for a further hour. Rabbit polyclonal antibodies directed to the 17 kd large subunit of caspase-3 (kindly provided by Dr D Nicholson, Merck-Frosst, Quebec, Canada), which recognize the proform of caspase-3 and its kd large subunit(s), and to the 32

43 Chapter 2 Materials and Methods carboxy terminus of caspase-2, which recognize both pro-caspase-2 and the 12 kd subunits, were used (Santa Cruz Biotechnology, CA). A mouse monoclonal antibody to poly (ADP-ribose) polymerase (PARP) was used, which recognizes both intact PARP (116 kd) and a cleavage product of 89 kd (kindly provided by Dr G Poirier, Laval University, Quebec, Canada). A polyclonal antibody to caspase-7 (kindly provided by X-M Sun, MRC Toxicology Unit, University of Leicester, Leicester, UK), which recognizes the pro-caspase-7 and the catalytically active 19 kd large subunit, was also used as described. A rabbit polyclonal antibody to caspase-8 was raised against the large subunit of caspase-8 (amino acids ) and kindly supplied by Dr X-M Sun (MRC Toxicology Unit, University of Leicester, Leicester, UK). The antibody obtained was characterized by ELISA and Western blot analysis, which verified that it recognized intact procaspase-8 and the 43 kd and 18 kd subunits. A rabbit polyclonal antibody to c-flip (kindly provided by Dr. D Nicholson, Merck-Frosst) which recognises the intact form of c-f L IP l and the intact and clipped forms of c-flips (Rasper et al, 1998), and a monoclonal antibody against FADD which recognises a 26 kd band corresponding to FADD in Jurkat cell lysates (BD Transduction Labs, Kentucky, USA) were also used. All antibodies were checked for specificity using lysates prepared from control cell lines. A sample of apoptotic cells from a cell line was also included on each blot of experimental samples. The cell lines used were THP-1 (human monocytic) and Jurkat (human T lymphoblastoid) treated with staurosporine (0.2 pm) for 4 h in order to induce caspase activation/substrate cleavage. The blots were developed using enhanced chemiluminescence (ECL, Amersham, Bucks, UK) according to the manufacturer s instructions. Blots were exposed to X-OMAT LS film (Kodak, Rochester, NY) for appropriate time intervals, and were developed using a Kodak X-OMAT film processor (Kodak, Rochester, NY), in order to visualize caspase activation / substrate cleavage. 33

44 Chapter 2 Materials and Methods 2.10 Stimulation of B-CLL cells using anti-cd40 monoclonal antibody and interleukin-4 Culture flasks were prepared as described by Walker et al (1997), by incubating overnight with goat-anti-mouse immunoglobulins (Dako, High Wycombe, Bucks) in TBS at a concentration of 5 pg/ml. 3.4 ml of this solution was used per 25cm3 volume of flask. Flasks were blocked for lh at room temperature using a solution of TBS/1% FCS. B-CLL cells were resuspended in culture medium (as previously described) and dispensed into the prepared flasks. Anti-human CD40 monoclonal antibody (R&D systems, Abingdon, UK) was added to a final concentration of 1.5 pg/ml. The cells were cultured overnight at 37 C. For stimulation with interleukin-4, cells were dispensed into culture flasks, and the media was supplemented with recombinant human interleukin-4 (R&D Systems) at a concentration of 10 ng/ml. Cells were then cultured for the time periods indicated in the text Analysis of CD95/Fas receptor expression on B-CLL cells 1 x 106 cells were resuspended in 100 pi of PBS/1 % FCS. 1 pi of anti-human Fas monoclonal antibody (CH11, Immunotech, Marseilles) and 5 pi of goat-anti-mouse- FITC antibody (Dako) were added. As a negative control, 1 x 106 cells were incubated with 1 pi control (non-human specific) IgM antibody (Dako) and 5 pi goat-antimouse-fitc. The cells were incubated on ice for 30 minutes prior to centrifugation at 6,000 rpm for 2 minutes. The cells were resuspended in 1 ml PBS for analysis by flow cytometry. Histogram plots of FL1-H (FITC) staining were obtained, the photomultiplier tube (PMT) voltage settings being determined by the level of staining in the negative control sample. 34

45 Chapter 2 Materials and Methods 2.12 Isolation of the death inducing signalling complex (DISC) Technique courtesy of Dr. M.E. Peter, German Cancer Research Center, Heidelberg, Germany. 1 x 10 SKW6.4 cells (murine B lymphoblastoid) were resuspended in 5 ml cell culture media (RPMI 1640, 10% FCS, 1% penicillin/streptomycin, 1% non-essential amino acids (GibcoBRL)). Anti-Apo-1 (IgG (Bender MedSystems, Vienna, Austria)) was added at 1 pg/ml, and the cells were cultured for 10 minutes at 37 C. Following stimulation, the cells were washed in ice-cold PBS and resuspended in 1 ml ice-cold lysis buffer (30 mm Tris-HCl (ph 7.5), 150 mm NaCl, 1% Triton-X 100, 10% glycerol, 1 pg/ml each aprotinin, leupeptin and pepstatin A). As negative controls, two aliquots of unstimulated cells were also lysed. The lysate from one of these samples was supplemented with 1 pg/ml anti-apo-1. The samples were vortexed and incubated on ice for 15 minutes. The samples were then spun at 13,000 rpm at 4 C for 15 minutes. 30 pi of a 50% solution of Protein-A Sepharose CL-4B (Sigma) was added to each lysate. The samples were then incubated at 4 C with rolling action for 1 h 30 mins. Following this incubation period, the samples were spun at 6,500rpm for 2 minutes, the supernatant removed, and the sepharose washed in 4 changes of lysis buffer. Sample loading buffer containing 10% SDS, 1% p-mercaptoethanol and bromophenol blue was then added, the samples were boiled and SDS-PAGE and immunoblotting was performed as previously described. A sample of whole Jurkat cells which had been incubated with anti-fas monoclonal antibody (clone CH11, Immunotech, Marseille, France) for 4 h was included on the gels as a control for antibody specificity. The membranes were probed with antibodies against FADD and caspase-8 (all as previously described). The experiment was repeated using Jurkat cells and B-CLL lymphocytes, where the initial stimulation period was increased to 60 minutes. 35

46 Chapter Three Results 1 Chapter 3 Development of an in vitro apoptosis sensitivity assay for CLL cells 3.1 Introduction For analysing the effects of chemotherapeutic drugs on ex vivo cells, toxicity assays such as the differential staining cytotoxicity assay (Bosanquet et al, 1993) and the MTT assay (Campling et al, 1988) have traditionally been used. Apoptosis has been shown to be the primary form of cell death induced by chemotherapeutic drugs. Many techniques for observing and quantifying apoptosis have been developed, which recognise various features of an apoptotic cell. The aim of this study was to develop an in vitro apoptosis sensitivity assay to measure susceptibility of B-CLL cells to the chemotherapeutic drugs prednisolone and chlorambucil. Initial experiments in this project used as a measure of apoptosis the extent of DNA fragmentation in the cells, using agarose gel electrophoresis to assess nucleosomal DNA cleavage, and the quantitative technique of in situ end labelling (ISEL), where analysis is performed flow cytometrically. As the study progressed and the need for an earlier marker of apoptosis became evident, the Annexin V assay was employed which measures the extent of extemalisation of phosphatidylserine. 3.2 CLL cells undergo spontaneous apoptosis in in vitro culture - measurement using the ISEL method In order to make a valid estimation of the in vivo apoptosis sensitivity of a malignant cell type in an in vitro system, it must be taken into account that removal of the malignant cells from their normal cellular environment may alter their innate susceptibility to induction of apoptosis. This most likely occurs due to the loss of one or more survival stimuli which are present in vivo, but which are not present in the cell culture environment. When estimating the sensitivity of ex vivo cells to chemotherapeutic drugs, this altered sensitivity must be taken into account. The term spontaneous apoptosis refers to the apoptosis level induced following culture of the CLL cells an in vitro culture system, without the addition of any apoptosis-inducing 36

47 Chapter Three Results 1 stimuli such as chemotherapeutic drugs. The existence of spontaneous apoptosis has been reported previously (Collins et al, 1989) and was evident from the first few experiments performed during this study, and will be referred to throughout this thesis. In order to determine whether cells from different patients were sensitive to spontaneous apoptosis to varying degrees, samples from nine patients were analysed as described below. The total lymphocyte fraction was purified from blood samples obtained from nine patients with B-CLL. The cells were labelled using the ISEL method to measure DNA fragmentation and analysed flow cytometrically (figure 3.1 A). In all nine patient samples, the level of apoptosis in the freshly isolated cells was less than 3% indicating a low rate of apoptotic cell death in vivo (figure 3.1 B). Lymphocytes purified from the same B-CLL patients were also cultured for 24 h. Subsequently, an aliquot of cells (1 x 106 cells) from each culture was labelled using the ISEL method and analysed flow cytometrically to measure the level of apoptosis in the culture. These cells analysed following culture in vitro without the addition of anti-cancer drugs showed a wide variation in spontaneous apoptosis (figure 3.1 B), indicating variable sensitivity to apoptosis induction between B-CLL cells from individual patients. 37

48 Chapter Three Results 1 Freshly isolated cells Control cells at 24 h O FL2-Area (6) v s FL1 -Height (3) o FL2-Area (6) v s FL1 -Height (3) <u X 3% / ' k 93% Pm 97% "I 1T 1 1I % B FL2-Area 30 _ 25 -I LU 2 20 <0 o a a. 10 < 24 h * I 4 J _ J J J _ I 121 Patient Number Figure 3.1 CLL cells were isolated from whole blood samples, and were labelled using the ISEL method and analysed flow cytometrically. Samples of the cells were also cultured for 24 h in vitro prior to being labelled and analysed using ISEL. A- Dot plots showing typical of results obtained from use of the ISEL method on freshly isolated and cultured CLL cells. On the X-axis, FL2-Area is a measure of the DNA content of the cells, and on the Y-axis, FL1-Height is a measure of green (FITC) flourescence. Cells expressing high green fluorescence are apoptotic. B- Analysis of freshly isolated cells revealed a low ex vivo rate of apoptosis (0 h). After 24 h culture in vitro, the level of spontaneous apoptosis varied widely between patient samples. 38

49 Chapter Three Results Apoptotic CLL cells exhibit limited intermicleosomal DNA cleavage, but evidence of large DNA fragmentation can be observed In situ end labelling measures the extent of DNA fragmentation in apoptotic cells. In order to visualise the extent of this DNA fragmentation in apoptotic CLL cells, agarose gel DNA fragmentation analysis was performed on freshly isolated CLL cells, and those which had been cultured in the presence of chemotherapeutic drugs. Following the culture of CLL lymphocytes for 24 h alone or in the presence of prednisolone ( mm), chlorambucil (15 pm), fludarabine (3 pm) or etoposide (2 pm), samples of cells were analysed by agarose gel electrophoresis for evidence of intemucleosomal DNA fragmentation. In all of the patient samples, after 24 h of culture, there was no clear evidence of intemucleosomal DNA cleavage (figure 3.2 A). Samples from five of the patients were analysed using field inversion gel electrophoresis (FIGE). This analysis revealed the presence of large fragments of DNA of 300 kb and 50 kb produced following induction of apoptosis by chemotherapeutic dmgs (figure 3.2 B). From the agarose gel electrophoresis analysis it appears that within 24 h of removal from the in vivo environment, with or without treatment with chemotherapeutic dmgs, CLL cells do not exhibit significant amounts of nucleosomal DNA cleavage, but do undergo cleavage of the DNA into large ( kb) fragments. This finding raised significant questions as to the accuracy of using DNA fragmentation as a measure of apoptosis in CLL cells. It was decided to employ an alternative flow cytometric method, the Annexin V assay, which measures the extemalisation of phosphatidylserine, a common feature of apoptotic cells (Koopman et al, 1994) 39

50 Chapter Three Results 1 4 h 24 h ( \ ( 'i M M mm Pd B < 700 kb 300 k B > 50kB> Figure 3.2 B-CLL lymphocytes were analysed by agarose gel electrophoretic methods in order to determine the extent of DNA fragmentation occuring during apoptosis in these cells. A. Cells were analysed for nucleosomal DNA cleavage after 4h and 24h culture alone or with prednisolone (Pd). B. Cells were analysed for formation of large (>50kB) DNA fragments in response to culture alone (lanes 2,3,c,d) or with prednisolone (10 mm, lanes 4 and 5; 1 mm, lanes 6 and 7; 0.1 mm, lanes 8 and 9) or chlorambucil (15 pm, lanes e and f) or etoposide (2 pm, lanes 10 and 11). (Lanes 1 and a are lambda phage genome concatamers, and lane b is S. cerevisiae chromosome marker). 40

51 Chapter Three Results The ISEL assay underestimates the percentage of apoptotic CLL cells In situ end labelling measures the extent of apoptotic DNA fragmentation in a population of cells. Previous results had demonstrated that CLL cells undergoing apoptosis exhibited only a limited amount of intemucleosomal DNA fragmentation (section 3.3), and so an alternative method of assessing apoptosis was chosen, the Annexin V assay. In the hope that results obtained using ISEL would relate to results gained using the Annexin V assay, comparative experiments were performed to establish the relationship between results from the two techniques CLL cells from six patients were cultured for 24 h alone (control) or in the presence of prednisolone (20 pm) or chlorambucil (7 pm) prior to being analysed using both the Annexin V and ISEL labelling techniques. The results obtained from use of the ISEL method were plotted against the results obtained from using the Annexin V assay (figure 3.4). Regression analysis confirmed that the two sets of data had a linear relationship which was highly significant (P < ) at the 99.5% confidence level. The equation for the regression line was calculated (y = 2.0lx), the coefficient of 2.01 indicating that the Annexin V values are double those recorded using the ISEL method. As a result of this finding, it was deemed unsuitable to continue using DNA fragmentation as a marker for apoptosis in CLL cells, and the Annexin V assay was used in all subsequent analysis in this thesis. 41

52 Chapter Three Results 1 80 y = x a o < 50 c ^ ISEL % Apoptosis Figure 3.4 A comparison of the ISEL and Annexin V labelling techniques demonstrates that measuring DNA fragmentation using ISEL underestimates the number of apoptotic CLL cells in a population by approximately 50 %. Cells were isolated from six CLL patients. At 0 h and after 24 h of culture alone or in the presence of 20 pm prednisolone or 7 pm chlorambucil, duplicate samples were labelled using ISEL and the Annexin V assay. The flow cytometry results obtained using each technique were plotted against each other. Regression analysis was performed to check for a linear relationship, and the regression line was plotted. The equation for the line is stated on the graph. 42

53 Chapter Three Results In vitro sensitivity to spontaneous apoptosis is a predictor of in vitro sensitivity to chlorambucil-induced apoptosis Previous results had demonstrated the variability between patient samples in spontaneous apoptosis sensitivity. To determine whether this sensitivity was reflected in the response of the cells to chemotherapeutic drug-induced apoptosis, CLL cells from 30 patients were tested for chlorambucil sensitivity. For each case, duplicate cultures were set up. One of these cultures was left untreated (control), and the chlorambucil was added to the second culture (7 pm). The cells were incubated at 37 C for 24 h. Apoptosis sensitivity in the two cultures was assessed flow cytometrically using the Annexin V assay to measure externalised phosphatidylserine. Control cultures were scored as sensitive or resistant to spontaneous apoptosis. (Sensitive samples were given the value 1, and resistant values were scored 0). The value chosen to delineate between the two groups was the median value of % spontaneous apoptosis. The chlorambucil-treated samples were then split into two groups based on the sensitivity to spontaneous apoptosis of their corresponding control cultures and the results were plotted (figure 3.5). A one-tailed T-test (assuming unequal variances) was performed in order to determine if the means of the two groups were significantly different. A significant P-value of < confirmed that the patients which were scored as sensitive to spontaneous apoptosis were most likely to be highly sensitive to chlorambucil-induced apoptosis. This implies that the relationship between sensitivity to spontaneous apoptosis and sensitivity to in vitro drug-induced apoptosis is significant, a finding which would greatly simplify any predictive testing technique based upon apoptosis sensitivity. 43

54 Chapter Three Results 1 80 ( CO 50 g 40 T l ^ E 20 6^ 10 t I 0 1 Sensitivity to spontaneous apoptosis Figure 3.5 Cells from thirty CLL patients were cultured for 24 h alone (control) in order to determine sensitivity to spontaneous apoptosis, or in the presence of 7 pm chlorambucil. After 24 h, flow cytometric analysis of the cells was performed using the Annexin V labelling technique to measure externalised phoshatidylserine. The control cultures were scored as sensitive (1) (n = 15) or resistant (0) (n = 15) to spontaneous apoptosis. The value obtained for chlorambucil-induced apoptosis for each sample was plotted against the spontaneous apoptosis score in each case. 44

55 Chapter Three Results As patients undergo chlorambucil therapy, the in vivo level of apoptosis can decrease, but the sensitivity of the cells to spontaneous and chlorambucil-induced apoptosis in in vitro culture increases In order to attempt to determine how closely the in vitro system of apoptosis induction by chemotherapeutic drugs was related to the in vivo reaction of CLL cells to drug therapy, a small in vivo study was performed. Blood samples were taken from two CLL patients (8 A and 12 1) immediately prior to them beginning a 14-day course of chlorambucil. This was to be the first course of treatment for patient 8 A, but patient 12 I had received chlorambucil therapy previously. In addition to the pre-treatment blood sample, further samples were taken from the same patients at day 7 of the course of treatment, and at day 30, two weeks post-treatment. On each sample day, the in vivo level of apoptosis was analysed by labelling the freshly isolated lymphocytes using the Annexin V technique. Flow cytometric analysis was performed immediately. This method attempts to give as accurate a measure of in vivo apoptosis rate as is possible, the time lapse from phlebotomy to flow cytometric analysis being less than 60 minutes. This analysis revealed that the in vivo apoptosis rate for patient 8 A remained relatively stable, whereas the in vivo apoptosis level for patient 12 I decreased as the course of treatment was administered (figure 3.6). A proportion of cells from each isolate were resuspended in culture medium and cultured for 24 h. Assessment of the level of spontaneous apoptosis induced in the cultures was made usisng the Annexin V assay. This analysis revealed that the in vitro apoptosis sensitivity of the ex vivo CLL cells was elevated during administration of chemotherapy. In both cases, the level of spontaneous apoptosis increased during and after the course of treatment when compared with the level of spontaneous apoptosis obtained pre-treatment (figure 3.6). In order to further compare the in vivo and in vitro responses of CLL cells to chlorambucil, the remaining cells from each sample were cultured in vitro in the presence of 7 pm chlorambucil. After 24 h culture, the sensitivity of these cells to chlorambucil-induced apoptosis was assessed flow cytometrically using the Annexin V technique. The cells from patient 12 I showed a decrease in sensitivity to 45

56 Chapter Three Results 1 chlorambucil as treatment commenced, which increased again by the 30 day sample point (figure 3.6). However, the sensitivity of the cells did not recover to the levels observed prior to the treatment beginning, possibly indicating the loss of a particularly chlorambucil-sensitive clone. Since this patient had in the past had one course of chlorambucil therapy, this result may indicate some degree of resistance of the remaining clone to apoptosis induction by chlorambucil. Additionally, the level of apoptosis induced by chlorambucil at the 30 day sample point was not significantly higher than the level of spontaneous apoptosis induced at the same timepoint. When compared with the greater difference in the cells sensitivity to spontaneous and chlorambucil-induced apoptosis at the pre-treatment sample point, this may indicate that the cells are becoming increasingly resistant to chlorambucil as a result of the therapy. Cells from patient 8 A showed a steady increase in sensitivity to chlorambucil as the course of treatment progressed, mirroring the increase in sensitivity to spontaneous apoptosis. Of interest is the observation that the post-treatment level of chlorambucilinduced apoptosis in this patient was 37% higher than the pre-treatment level (figure 3.6), which implies that a number of chlorambucil-sensitive CLL cells remain in circulation following the end of the course of treatment. 46

57 Chapter Three Results C/5 o co 60 Q. O Q. < Patient 12 I Freshly isolated cells Control 7 um Chlorambucil Re-treatment (Day 0) Day 7 Sample Day Fbst-treatment (Day 30) 100 Patient 8 A <0 '(0 2Q. O Q. < Re-treatment (Day 0) Day 7 Sample Day Fbst-treatment (Day 30) Figure 3.6 In vivo apoptosis can decrease and sensitivity to spontaneous apoptosis can increase as drug treatment is administered. Blood samples were taken from patients 121 and 8 A prior to them starting a two week course of chlorambucil (day 0). Blood samples were also taken at day 7 of their courses of treatment, and at day 30 (post-treatment). At each sample point, freshly isolated lymphocytes were labelled and analysed using the Annexin V technique. Other cells from the same isolate were cultured for 24 h in culture medium alone (control) or with 7 pm chlorambucil before being analysed using the Annexin V assay. For flow cytometry histograms of freshly isolated and control cells at each of the sample points for both patients, please refer to the Appendix, figure A1. 47

58 Chapter Three Results Discussion The initial aim of this reseach project was to investigate the relationship between chemotherapy and apoptosis in chronic lymphocytic leukaemia (CLL). CLL is often diagnosed in an asymptomatic phase, and patients can live with the disease for several years before in some cases the tumour burden and its accompanying effects on the body mean that chemotherapeutic treatment is necessary. The treatment of choice has been the alkylating agent, chlorambucil with or without the glucocorticoid, prednisolone. More recently, the purine analogues, of which fludarabine is one, have shown increasing promise and a higher rate of remissions when used as first line therapy, when compared with the remission rate obtained using chlorambucil. However, drug resistance is a major problem in CLL, and second or third treatments with chlorambucil do not usually produce high remission rates. Since the advent of chemotherapy, it has been a goal for many researchers to develop a means of analysing the toxicity or killing efficacy of the drugs on the target cells in order to reduce the incidence of unsuccessful treatment (Hanson et al, 1991; Bosanquet, 1993). In addition the effect of the drugs on other cells in the body needs to monitored, a process which plays a crucial role in the development and testing of new drug therapies. Toxicity tests have been in existence for many years, and have been used extensively for purposes such as those outlined above. Examples include the MTT assay which assesses the amount of insoluble formazan produced by living cells (Campling et al, 1988) and the differential staining cytotoxicity assay (Bosanquet et al, 1983). In the 1970 s apoptosis was identified as a mode of cell death distinct from necrosis (Kerr, 1972). Since then, investigations into the effects of chemotherapy on mammalian cells have revealed that apoptosis is the primary method by which cells die as a result of exposure to these drugs. This study began with the development of a method to analyse apoptosis sensitivity of lymphocytes sourced from CLL patients attending outpatient clinics, some untreated, and some who were receiving chemotherapy. At the time that this study commenced there were numerous techniques available for analysing or quantifying apoptotic cells (Carbonari et al, 1995), several of which were based upon the characteristic 48

59 Chapter Three Results 1 intemucleosomal DNA cleavage that occurs in an apoptotic cell. Qualitative agarose gel electrophoresis techniques and flow cytometric methods, such as in situ end labelling, appeared to be the ideal techniques to apply to CLL samples. In particular, in situ end labelling (ISEL) (Wolfe et al, 1996) was chosen over some newer flow cytometric labelling techniques (such as the Hoerscht high blue, method (Brown et al, 1996)), because of the inclusion of a fixation step in the procedure. This enabled the samples to be stored until a sufficient number were available to be analysed, and also decreased the risks asscociated with handling unfixed clinical specimens. The data obtained from the flow cytometric analysis of the CLL cells were supplemented using agarose gel electrophoretic methods to enable visualisation of intemucleosomal and larger DNA fragmentation (Figure 3.2 A and B). Initial results using ISEL analysis revealed a low level of in vivo apoptosis in the cells from every patient analysed, demonstrating that the CLL lymphocytes are not undergoing cell death in the patient to any significant degree (Figure 3.1 B). It is possible that the in vivo level could be higher than this, but that apoptotic cells are being cleared effectively from the blood, and so cannot be measured as such. However, if these figures do accurately represent levels of in vivo apoptosis, then this supports the lymphoaccumulative model of CLL (Dameshek, 1967), in which CLL lymphocytes remain for a long period in circulation due to reduced levels of cell death. ISEL analysis also confirmed the existence of spontaneous apoptosis, a feature of cultured ex vivo CLL cells which had been observed previously (Collins et al, 1989). From a common low level of in vivo apoptosis, CLL cells cultured for 24 h prior to being labelled using the ISEL method and analysed flow cytometrically varied dramatically in their sensitivity to spontaneous apoptosis (Figure 3.1 B). This implies that differences may exist between the CLL lymphocytes of the spontaneous apoptosis sensitive and resistant patients. Differences in the expression levels of apoptosiscontrolling factors could be one reason for this variability. The relative levels of antiapoptotic Bcl-2 and pro-apoptotic Bax in these cells may be instrumental in determining their apoptosis sensitivity, a theory which has been investigated by a number of research groups (McConkey et al, 1996; Gottardi et al, 1996; Pepper et al, 49

60 Chapter Three Results ; Thomas et al, 1996). A high Bcl-2:Bax ratio has been shown to correlate with resistance to chlorambucil in vitro (Thomas et al, 1996), and Bax upregulation appears to be required for chemotherapeutic drug-induced apoptosis to take place, whilst a high expression level of Bcl-2 confers survival (Pepper et al, 1999). It is also likely that CLL cells differ in their sensitivity to growth factors in their surrounding environment, a subject which will be discussed in later sections of this thesis. Populations of cells which depend heavily on growth factor stimulation (such as less mature and less well differentiated CLL clones) are likely to be more susceptible to cell death upon growth factor withdrawal (such as occurs in in vitro culture). Use of gel electrophoretic methods to confirm that apoptosis was indeed the prevailing mode of cell death in CLL cells exposed to chemotherapeutic drugs led to the observation that, following 24 h of in vitro culture, there was little evidence of the intemucleosomal cleavage characteristic of apoptosis (Figure 3.2 A). However, large DNA fragmentation analysis demonstrated that the cells were in the early stages of DNA fragmentation confirming that apoptosis was occuring (Figures 3.2 B and 3.3). This phenomenon of little or no intemucleosomal cleavage in apoptotic CLL cells had been reported previously (Huang & Plunkett, 1995), but the presence of intemucleosomal DNA fragments may be dependent on the sampling point chosen following apoptosis induction. Later time points may pick up DNA cleavage more effectively than earlier time points. However, since the ISEL technique is based upon labelling the fragmented DNA with digoxygenin-labelled nucleotides in order to visualise the apoptotic cells, it was of some concern that the cells in question were not undergoing intemucleosomal DNA fragmentation within the timescale of the assay. In addition, since the study was moving towards more in-depth analysis of apoptosis in CLL cells, it became more pmdent to use an earlier marker than DNA fragmentation in order to quantify apoptosis in these cells. Annexin V has a high affinity for the membrane phospholipid, phosphatidylserine. During apoptosis, phosphatidylserine (PS) is flipped from the inner leaflet of the cell membrane to the outer surface, where it serves as a recognition marker to phogocytic cells (Koopman et al, 1994), and where it can also be bound by fluorescently labelled recombinant Annexin V and used as a marker for flow cytometric analysis of 50

61 Chapter Three Results 1 apoptosis (Martin et al, 1995). In order to introduce this new technique into the study, a small series of comparitive experiments were performed. Cells from six CLL patients were cultured alone or for 24 h in the presence of chemotherapeutic drugs prior to being labelled and analysed using both the ISEL and Annexin V techniques. When the results obtained with each technique were plotted against the other, it was discovered that the proportion of apoptotic cells in the CLL cell population was being underestimated by ISEL by as much as 50%, confirming the observation that DNA fragmentation was not taking place in these cells to any great degree (Figure 3.4). All further analysis o f apoptosis levels in this study was made using the Annexin V assay. Large variation had been observed in the sensitivity of CLL cells from different patients to spontaneous apoptosis. To determine if this sensitivity to spontaneous apoptosis was mirrored in the response of the cells to chlorambucil-induced apoptosis, cells from thirty patients were cultured alone or in the presence of chlorambucil for 24 h, prior to assessment of apoptosis using the Annexin V assay. When the results were compared, it was discovered that cells which were sensitive to spontaneous apoptosis were also more likely to be sensitive to apoptosis induced by chlorambucil (p < ). This implies that there is a relationship between sensitivity to spontaneous apoptosis and sensitivity to in vitro drug-induced apoptosis, a finding which could greatly simplify predictive testing techniques based upon apoptosis sensitivity. In order to assess how closely this in vitro analysis of CLL cells response to chemotherapeutic agents was related to the in vivo response to chemotherapy, and to analyse how the apoptosis sensitivity of the cells altered during a course of treatment, a small in vivo study was performed. Cells were taken from two CLL patients prior to, during and after they had received a 2-week course of chlorambucil. As the treatment commenced the in vivo level of apoptosis in one of the patients decreased, whilst the level in the other patient remained stable. More interesting was the effect on the in vitro spontaneous apoptosis sensitivity of the cells. As the patients course of treatment progressed, the CLL cells became increasingly sensitive to spontaneous apoptosis, possibly indicating that the treatment was having the desired effect, in that the cells were being triggered into apoptosis, removal to the in vitro situation serving only to hasten their eventual death. The fact that the cells were more sensitive to 51

62 Chapter Three Results 1 spontaneous apoptosis two weeks after the end of the course of treatment than they were prior to the treatment commencing may have some bearing on treatment decisions for future patients. However, a much larger group of patients would need to be analysed using this method before this observation could be confirmed. Samples of cells taken from the treated patients were also cultured for 24 h in the presence of 7 pm chlorambucil. Analysis of the chlorambucil sensitivity of these cells using the Annexin V assay revealed differences between the two patients studied. Patient 12 I had previously undergone treatment with chlorambucil. The cells from this patient decreased in their sensitivity to in vitro chlorambucil-induced apoptosis as the course of therapy progressed. Two weeks post-treatment, the cells had recovered some sensitivity to chlorambucil-induced apoptosis, but not to the level seen at the pre-treatment sample point. This could signify the loss of a chlorambucil-sensitive clone of cells and may indicate some degree of resistance of the remaining clone to apoptosis induction by chlorambucil. Since the patient had been treated with this drug in the past, it may be conceivable that some degree of chlorambucil-resistance had developed. In support of this theory, increases in p-glycoprotein (Perri et al, 1989) and upregulation of glutathione-s-transferase mrna expression (Schisselbauer et al, 1990) have been demonstrated in some cases of chlorambucil-resistant CLL. The cells from patient 8 A behaved in a slightly different manner. For this patient, this was the first course of treatment with chlorambucil or any other chemotherapy. The cells from this patient showed a steady increase in sensitivity to apoptosis induction in vitro by chlorambucil as the course of treatment progressed. Of some interest is the observation that the post-treatment level of chlorambucil-induced apoptosis in this patient was higher than the pre-treatment level. This may imply that a significant number of chlorambucil-sensitive CLL cells remain in circulation following the end of the course of treatment, a finding which also may have relevance to the way in which CLL patients are treated therapeutically. Again, a larger group of previously treated and untreated patients would need to be analysed before this observation could be confirmed. 52

63 Chapter Four Results 2 Chapter 4 - Processing/Activation of caspases -3, -7, and -8, but not caspase-2 in the induction of apoptosis in B-chronic lymphocytic leukaemia cells 4.1 Introduction The importance of caspase activation in spontaneous and chemotherapy-induced apoptosis in leukaemic cells from patients with CLL had not previously been described. It had been shown that peripheral blood mononuclear cells from CLL patients were caspase-3 immunopositive (Krajewski et al, 1997) and that glucocorticoid-induced apoptosis of CLL lymphocytes requires protease activation and is accompanied by cleavage of PARP and lamin Bi, together with loss of caspase- 3 (Bellosillo et al, 1997; Chandra et al, 1997). The aim of the work described in this chapter, was to further define the role of caspases in induction of spontaneous and drug-induced apoptosis in CLL cells. 4.2 Inhibition of spontaneous apoptosis in CLL cells by Z-VAD.fmk Apoptosis was assessed in CLL cells using Annexin V to measure extemalisation of phosphatidylserine (figure 4.1). Freshly isolated cells, prior to culture, showed a very low level of spontaneous apoptosis, as measured by Annexin V (figure 4.1 A). During culture, the cells underwent spontaneous apoptosis (figure 4.IB), which was increased in the presence of chemotherapeutic agents, such as chlorambucil (figure 4.1C). A variable amount of spontaneous apoptosis was observed (Tables 4.1 and 4.2) in agreement with other studies (Collins et al, 1989). The measurement of phosphatidylserine exposure allowed a clear quantification of the percentage of apoptotic cells. In order to assess the role of caspases in the execution phase of apoptosis in B-CLL cells, Z-VAD.fmk, a cell permeable caspase inhibitor, was used which inhibits apoptosis in many different model systems (MacFarlane et al, 1997, Zhu et al, 1995, Feamhead et al, 1995). Z-VAD.fmk (100 jlim ) inhibited spontaneous apoptosis in all but one of the patient samples examined (Table 4.1) supporting the involvement of caspases in the spontaneous apoptosis of CLL cells. In order to obtain more definitive evidence for the involvement of caspases, the processing/activation of the effector caspases was examined as well as the proteolysis of PARP, which has been used as a marker of apoptosis ( Kaufmann et al, 1993). 53

64 Chapter Four Results % 93.6% CO 66.2% M2 33.9% M % M % 16A 4 Annexin V - FITC 10* J 10* Figure 4.1 Induction of apoptosis in cells from two patients with CLL assessed by phosphatidylserine extemalisation. A - Freshly isolated CLL cells, from patients 15 A and 16 A, were examined for phosphatidylserine exposure by binding of Annexin V as described in Chapter 2, section 2.6. The percentages of cells with low and high Annexin V binding representing normal (marker 1) and apoptotic (marker 2) cells are shown. CLL cells were also cultured for 20 h either B. alone to measure spontaneous apoptosis or C. in the presence of chlorambucil (7.5 pm). 54

65 Chapter Four Results 2 Patient No % Spontaneous apoptosis - Z-VAD.fmk + Z-VAD.fmk 2 A A A A A A 7 7 Table 4.1 Inhibition of spontaneous apoptosis by Z-VAD.fmk Freshly isolated CLL cells from patients were cultured for 24 h at 37 C either alone or in the presence of Z-VAD.fmk (100 pm) as indicated. Apoptosis was assessed by measuring extemalisation of phosphatidylserine using Annexin V staining. 55

66 Chapter Four Results Activation of caspase-3 and caspase-7 in apoptosis of CLL cells. The time course of induction of apoptosis was examined in four previously untreated Binet Stage A cases of B-CLL. In each patient, a time dependent induction of spontaneous apoptosis was observed (figure 4.2 A). Cells from these patients exposed to chlorambucil, showed a concentration dependent increase in the induction of apoptosis compared to control cells (figure 4.2 A). These cells were also analysed by immunoblotting for activation of caspases -3 and -7. Freshly isolated untreated CLL cells contained primarily the proform of caspase-3 (figure 4.2 B, lane 1). Processing of caspase-3 at Asp 175 between the large and small subunits yields a 20 kd subunit, which is further processed at Asp 9 and Asp 28 to yield 19 kd and 17 kd large subunits (Fernandez-Alnemri et al, 1996). A time dependent processing of caspase-3 to its catalytically active large subunit(s), kd, was observed (figure 4.2 B, lanes 2-5), which was increased in the presence of chlorambucil (figure 4.2 B, lanes 6-9). Activation of caspase-3 was observed in both spontaneous and drug-induced apoptosis in all 9 samples examined to date (Table 4.2). Caspase-7 was also present in freshly isolated CLL cells as a ~35 kd proform without any detectable large 19 kd subunit (figure 4.2 C, lane 1). Processing of caspase-7 initially occurs at Asp 198 between the large and small subunits, followed by cleavage at Asp 23 to yield the 19 kd large subunit (MacFarlane et al, 1997). A time dependent processing of caspase-7 to its 19 kd subunit was observed in spontaneous apoptosis, which was also increased as a result of exposure to chlorambucil (figure 4.2 C). Caspase-7 was activated in all 10 samples of spontaneous and drug-induced apoptosis examined (Table 4.2). Thus induction of apoptosis in CLL cells was accompanied by the processing of both the effector caspases-3 and -7 to their catalytically active large subunits. 56

67 Chapter Four Results 2 (/> o + > a. o Time (h) B Control 7 pm Chi Time (h) 0 32 kd r N /" LS Control 7 pm Chi r " n r - \ Time (h) kd 19 kd xo l ' Figure 4.2 Induction o f apoptosis in cells from a representative CLL patient is accompanied by processing of caspase-3 and caspase-7. (A) Freshly isolated CLL cells were incubated either alone ( - ) or in the presence o f chlorambucil (3 pm) ( - ) or (7.5 pm) (A -A ) and apoptosis assessed at the indicated times by extemalisation o f phosphatidylserine as shown in Fig 4.1. (B) Processing of caspase-3 in CLL cells undergoing apoptosis. CLL cells from patient 15 A were cultured for the indicated times either alone (Con) or in the presence of chlorambucil (Chi, 7.5 pm) and examined by western blot analysis for caspase-3 as described in Chapter 2, section 2.9. The arrows denote either the 32 kd proform of caspase-3 or the catalytically active large subunits (LS) of approximately kd. (C) Processing of caspase-7 in CLL cells undergoing apoptosis. CLL cells from patient 15 A were cultured for the indicated times either alone (Con) or in the presence of chlorambucil (Chi, 7.5 pm) and examined by western blot analysis for caspase-7. The arrows denote either the 35 kd proform of caspase-7 or the large subunit of 19 kd. 57

68 Chapter Four Results 2 Patient No Sex/ Age Binet stage Previous therapy WBC x 109/1 % Apoptosis Spon Pd Chi -2 Caspase A M/74 A ND ND 3A M/72 A Chi ND ND 8A M/72 A Chi ND + ND ND ND 10A M/91 A ND ND 11A F/74 B Chi ND ND 12A F/67 B A M/62 A ND ND 14A F/60 A ND 42 ND + + ND + 15A M/51 A ND 48 ND A M/81 A Chi ND 66 ND ND ND f* + 17A M/71 A ND ND + ND 18A F/84 B Chi, Pd ND + + ND 19A M/77 C Chi, Pd ND + + ND 20A F/85 A ND ND ND + ND PARP Table 4.2. Clinical information and summary of in vitro apoptosis sensitivity and incidence of caspase activation. Peripheral blood lymphocytes from patients diagnosed with B-CLL were cultured in vitro for 20 h either alone or in the presence of prednisolone (Pd) and chlorambucil (Chi). Both spontaneous (spon) and druginduced apoptosis were then determined by Annexin V binding. Activation of caspases -2, -3, -7 and -8 as well as proteolysis of PARP was determined by western blot analysis. Expression of the pro-forms of the caspases was found in all cases examined. A *+ indicates activation of the pro-enzyme to an active subunit was observed following in vitro culture with and without addition of the drugs, whereas a indicates no such activation was observed. (ND indicates not determined). 58

69 Chapter Four Results Caspase-2 processing does not generally accompany apoptosis of CLL cells. Several recent studies have shown that caspase-2 (ICH-1, Nedd2) is processed in some but not all cells during the induction of apoptosis (MacFarlane et al, 1997, Li et al, 1997, Harvey et al, 1997). The prodomain of caspase-2 binds to the adaptor molecule RAIDD/CRADD which also binds the receptor-interacting protein RIP and may thus regulate apoptosis (Duan & Dixit, 1997, Ahmad et al, 1997). It is not known which if any intracellular substrates are cleaved by caspase-2 in apoptosis. In order to investigate the possible importance of caspase-2 in apoptosis of CLL cells, an antibody was used which recognizes both the proform and the small p i2 subunit of caspase-2. In agreement with previous studies (MacFarlane et al, 1997), induction of apoptosis in the human monocytic tumor cell line, THP. 1, resulted in the activation of caspase-2 to its ~ 12 kd small subunit (figure 4.3, lane 1) and was included as a positive control for the processing of caspase-2. In cells undergoing spontaneous and drug-induced apoptosis, processing of procaspase-2 was seen in only 1/9 samples examined to date (Table 4.2). The results from this case are shown in more detail (figure 4.3). Untreated freshly isolated CLL cells from this patient contained primarily the ~ 48 kd proform of caspase-2 with no detectable small 12 kd subunit (figure 4.3, lane 2). These freshly isolated cells also contained an immunologically reactive protein of ~ 33 kd, which may represent an early processed form of caspase-2 (figure 4.3, lane 2). In cells undergoing spontaneous apoptosis, little or no cleavage of procaspase-2 to its 12 kd small subunit was observed (figure 4.3, lane 3). Induction of apoptosis with prednisolone resulted in the processing of procaspase-2 to its 12 kd subunit (figure 4.3, lane 4), which was more marked with chlorambucil (figure 4.3, lane 5). Thus although processing of caspase-2 was observed in cells from this patient, it did not appear to be a general feature accompanying either spontaneous or druginduced apoptosis of CLL cells. 59

70 Chapter Four Results 2 + Z - VAD.fink + ve Oh Con Pd Chi Con Pd Chi 48 kd^* 33 kd^* 12 kd Figure 4.3 Activation of caspase-2 only occurred in cells from one patient. Freshly isolated CLL cells, from patient 2A, were cultured for 20 h either alone (Con) or with prednisolone (Pd, 200 pm) or chlorambucil (Chi, 15 pm) in the absence or presence of the caspase inhibitor Z-VAD.fmk (100 pm) as indicated. Freshly isolated cells (0 h) were included as a control. The cells were examined by western blot analysis for caspase-2 as described in Chapter 2, section 2.9. The arrows indicate the proform of caspase-2 and its 12 kd small subunit. Of the 9 patients examined, this was the only one to show activation of procaspase-2. Human monocytic THP.l cells treated with etoposide (25 pm) for 4 h were included as a positive control (+ve) as caspase-2 is processed in these cells to its 12 kd small subunit. 60

71 Chapter Four Results Activation of the effector caspases result in the cleavage of PARP. As processing of caspases-3 and -7 was observed, CLL cells were also examined for the cleavage of PARP, a substrate for both these caspases (Nicholson et al, 1995, Fernandez-Alnemri, 1995). Induction of spontaneous apoptosis was accompanied by a time dependent cleavage of PARP to a characteristic 89 kd signature fragment (figure Patient 14A, lanes 2-5; Patient 17A, lanes 1-4). The induction of apoptosis and the cleavage of PARP were both induced following exposure of the CLL cells to chlorambucil (figure Patient 14A, lanes 6-9; Patient 16A, lanes 5-12), further supporting the involvement of caspases in the execution phase of apoptosis in CLL cells. Patient 14 A Time (h) 0 r Control 7 pm Chi a r kD > 89 k D > Patient 16 A Time (h) Control 3 pm Chi 7 pm Chi 1 r > r kD>" 8 9 k D > Figure 4.4 Cleavage o f PARP accompanies apoptosis in CLL cells. Freshly isolated CLL cells, from patients 14 A and 16 A, were cultured for the indicated times either alone (Con) or with chlorambucil (Chi, 3 pm or 7.5 pm) and then analysed by western blot analysis for intact PARP (116 kd) or its cleaved product (89 kd). 61

72 Chapter Four Results Activation of caspase-8 during apoptosis of CLL cells. Although caspase-8 has been implicated as one of the main activator caspases in receptor-mediated apoptosis (Boldin et al, 1996, Muzio et al, 1996), its role if any in drug-induced apoptosis is not clear. In order to examine its possible involvement in apoptosis of CLL cells, a caspase-8 antibody was utilised. In freshly isolated CLL cells from 7/7 patients, the antibody recognised two protein(s) of -55 kd (figure 4.6), most probably corresponding to two different isoforms of caspase-8, MACHal and MACHa2 (Boldin et al, 1996, Scaffidi et al, 1997). In addition the freshly isolated cells also contained two immunoreactive proteins of - 43 kd (figure 4.5 A and B, lane 1), which probably resulted from loss of the small 12 kd subunit following cleavage at Asp 374 (Medema et al, 1997). The cells also contained two immunoreactive proteins of -28 kd, which may have arisen following removal of the two death effector domains following cleavage at Asp 216. Culture of the CLL cells resulted in a time dependent processing of the -55 kd protein(s) to fragments of -43, -28 and a small amount of an 18 kd fragment (figure 4.5 A, lanes 2-5). Formation of the 18 kd large subunit most probably occurs following further cleavage of the -43 kd fragments at Asp 216. Formation of all these fragments was slightly increased following treatment of CLL cells with chlorambucil (figure 4.5 A, lanes 6-9). The time course of cleavage of caspase-8 appeared similar to that of caspase-3 and caspase-7. Some variation in the processing of caspase-8 was noted in the samples analysed to date. For example in patient 17A, caspase-8 was activated to fragments of -43 and -28 kd with little or no 18 kd being formed (figure 4.5 B). Thus interindividual variation was observed in the activation of caspase-8. Further investigation indicated that this interindividual variation in caspase-8 activation was apparent in a larger number of cases. Particularly striking in all samples was the apparently high levels of the proform of caspase-8 and the relatively small amounts of the proform which were processed during culture. 62

73 Chapter Four Results 2 Time (h) 0 2 Con Chi k D > 4 3 k D > 2 8 k D > 18kD> B 55kD> 43kD> + Z-VAD.fmk ( ^ 0 h Con Chi Pd Con Chi Pd + ve 28kD> 18kD> III Figure 4.5 Processing of caspase-8 in CLL cells is inhibited by Z-VAD.fmk. (A) CLL cells from patient 15 A were cultured for the indicated times either alone (Con) or in the presence of chlorambucil (Chi, 7.5 pm). (B) CLL cells from patient 17 A were cultured for 20 h either alone (Con) or in the presence of prednisolone (Pd, 20 pm) or chlorambucil (Chi, 7.5 pm) in the absence or presence of Z-VAD.fmk (100 pm) as indicated. Cells from both patients were analyzed for activation of caspase-8 by immunoblotting. Freshly isolated cells (0 h) were included as a control. In both cases THP.l cells induced to undergo apoptosis by exposure to staurosporine (0.5 pm) were included as a positive control (+ ve) for the processing of caspase-8. 63

74 Chapter Four Results Z-VAD.fmk inhibits the processing of caspases in CLL cells. In order to determine at what stage of the apoptotic process, Z-VAD.fmk was inhibiting apoptosis (Table 4.1), its ability to inhibit the processing of different caspases was examined. Freshly isolated CLL cells from patient 2A contained primarily the proform of caspase-3 (figure 4.6, lane 2). Spontaneous apoptosis was accompanied by the processing of caspase-3 to its catalytically active large subunit(s) (figure 4.6, lane 3), which was increased following treatment with both prednisolone and chlorambucil (figure 4.6, lanes 4 and 5 respectively), commensurate with their ability to induce apoptosis. In both spontaneous and drug-induced apoptosis, Z- VAD.fmk almost completely inhibited the processing of procaspase-3 to its catalytically active large subunits (figure 4.6, lanes 6-8). Z-VAD.fmk also inhibited the drug-induced processing of caspase-2 observed in patient 2A (figure 4.3, lanes 7 and 8). In patient 17A, Z-VAD.fmk inhibited the activation/processing of caspase-8 to a p43 fragment (figure 4.5 B, lanes 5-7). Thus Z-VAD.fmk acts to block drug-induced apoptosis of CLL cells by blocking the activation/processing of caspases. 64

75 Chapter Four Results 2 + Z - V A D.fm k A. ^ + ve Oh Con Pd Chi Con Pd Chi r w (mm 1 1^ ^ m m m m m m m m m LS Figure 4.6. Processing of caspase-3 in CLL cells is inhibited by the caspase inhibitor, Z-VAD.fmk. CLL cells from patient 8 I were cultured for 20 h either alone (Con) or in the presence of prednisolone (Pd, 200 pm) or chlorambucil (Chi, 15 pm) either in the absence or presence of Z-VAD.fmk (100 pm) as indicated. Freshly isolated cells (0 h) and THP.l cells treated with etoposide to induce apoptosis were included as controls. The cells were analyzed for activation of caspase-3 by immunoblotting. The arrows denote either the 32 kd proform of caspase-3 or the catalytically active large subunits (LS) of approximately kd. 65

76 Chapter Four Results Discussion In this study it has been demonstrated that induction of apoptosis of CLL cells leads to the selective induction of some but not all caspases. The processing/activation of at least three caspases (caspases-3, -7 and -8) has been shown during the execution phase of apoptosis of CLL lymphocytes whereas caspase-2 does not generally appear to be activated. Recent studies have implicated a role for the release of mitochondrial cytochrome c in the activation of procaspase-3 provided Apaf-3, now identified as caspase-9, and datp are present (Zou et al, 1997; Li et al, 1997). In all the patients examined, activation of the effector caspases, -3 and -7, were observed (Table 4.2 and figures 4.2 and 4.6) suggesting that these caspases may be responsible for the cleavage of PARP found in CLL cells in this and other studies (Bellosillo et al, 1997; Chandra et al, 1997). Further support for this hypothesis was provided by the observations that Z-VAD.fmk, a cell permeable caspase inhibitor, inhibited the activation/processing of all the caspases studied, as well as the cleavage of PARP and both spontaneous and drug-induced apoptosis of CLL cells (Table 4.1 and figures 4.2, 4.3, 4.5 and 4.6). Thus these effector caspases also play a central role in the execution phase of apoptosis in CLL cells as they do in other cell systems (Cohen, 1997; Nicholson & Thomberry, 1997). In 8/9 patients, caspase-2 was not activated (Table 4.2). In only one case was activation observed (figure 4.3). At the time of sampling this patient was Binet stage A, but in the last year has progressed to stage C. Activation of caspase-2 during the execution phase of apoptosis has been observed in different tumor cell lines (MacFarlane et al, 1997; Li et al, 1997; Harvey et al, 1997). The precise mechanism by which procaspase-2 is activated in cells is not known. It may involve recruitment through an adapter such as RAIDD/CRADD (Duan & Dixit, 1997; Ahmad et al, 1997) or based on activities of recombinant enzymes, it has been proposed that caspase-3 activates pro-caspase-2 (Li et al, 1997; Harvey et al, 1996). No activation of procaspase-2 was observed even in the presence of activated caspase-3 and caspase-7 demonstrating that neither caspase-3 nor -7 activates procaspase-2 in CLL cells. The data also demonstrates that activation of caspase-2 is not required for the induction of apoptosis in CLL cells. 66

77 Chapter Four Results 2 The results on the activation/processing of caspase-8 are particularly intriguing. Triggering of the Fas/CD95 receptor leads to the recruitment and activation of caspase-8, which may then act as the apical initiator caspase responsible for the activation of other caspases and the execution of apoptosis (Boldin et al, 1996; Muzio et al, 1996). Whilst caspase-8 is activated early in Fas/CD95-induced apoptosis, little is known as to its role if any in drug-induced apoptosis. Of particular interest were the high levels of procaspase-8 present in 7/7 patients (Table 4.2). Procaspase-8 was present as two main forms of 55 and 53 kd, which correspond to caspase-8/a (MACHal) and caspase-8/b (MACHa2) (Scaffidi et al, 1997). At the end of the culture, despite the induction of significant apoptosis (25-65 %), most of the caspase- 8 was still present as the proform (figure 4.5), in contrast to Fas/CD95-induced apoptosis when all the caspase-8 is rapidly processed (Scaffidi et al, 1997). In addition in 2/4 cases, caspase-8 was activated concurrently with caspases -3 and -7 after 6 to 10 h of in vitro culture. These results would seem to preclude an initiator role for caspase-8 in drug-induced apoptosis of CLL cells, and suggest that it may be activated non-specifically by other caspases. Some cytotoxic agents, such as doxorubicin, induce apoptosis in human leukaemic cells lines and neuroblastoma cells via the Fas/CD95 receptor/ligand system, a hypothesis supported by the finding that these cells also display cross-resistance between the cytotoxic agent and Fas/CD95-induced apoptosis (Freisen et al, 1996; Fulda et al, 1997; Freisen et al, 1997). However, other studies have suggested that chemotherapy-induced apoptosis is not dependent on Fas/CD95 receptor/ligand interaction (Eischen et al, 1997; Gamen et al, 1997). While the reasons for these discrepancies are not clear, they may be related to the rate of induction of apoptosis. For example, in those systems where a positive relationship between chemotherapy and involvement of the Fas/CD95 receptor/ligand system has been implicated (Freisen et al, 1996; Fulda et al, 1997; Freisen et al, 1997), apoptosis is induced over a long period of time (24-48 h) so allowing the synthesis of new proteins. The possible involvement of the Fas/CD95 receptor/ligand system in CLL cells remains to be determined. 67

78 Chapter Four Results 2 Interestingly CLL cells have been reported to have undetectable or very low levels of CD95 expression (Mapara et al, 1993; Moller et al, 1993; Wang et al, 1997). This expression can be increased following in vitro activation with interleukin-2, Staphlyococcus aureus 1 or CD40 although in some cases this may lead to proliferation (Marpara et al, 1993; Wang et al, 1997). Thus CLL cells express high levels of caspase-8 in conjunction with very low levels of CD95 receptor. These results highlight the possibility of developing new therapies for CLL based on the upregulation of CD95- or other death receptors, which would synergize with the high levels of caspase-8 in CLL cells. Freshly isolated B-CLL cells possess caspase-2, -3, - 7 and -8. Some of these caspases can be activated to cleave protein substrates such as PARP. These results demonstrate that B-CLL cells possess the complete apoptotic machinery required to execute the apoptotic programme. Thus the dysregulation of apoptosis in CLL cells does not appear to be due either to a deletion of pro-caspases or to point mutations leading to their inactivation. Rather, this study suggests that the molecular basis of dysregulated apoptosis in vivo may reside in the signalling leading to the activation of caspases, or the presence of inhibitory proteins at the apoptosis induction stage. 68

79 Chapter 5 Results 3 Chapter 5 - Studies on survival factors and the Fas signalling pathway in B-CLL 5.1 Introduction Spontaneous apoptosis occurs when B-CLL cells are removed from their normal environment and placed into in vitro culture (Collins et al, 1989 and section 3.2 of this thesis). This implies that one or more survival factors needed by the B-CLL cells are not being provided by the in vitro culture conditions. All previous analysis in this study had been performed on populations of mixed lymphocytes, both B and T cells. The T cell fraction can be assumed to be small due to the high proportion of tumour B cells, but still the presence of T lymphocytes may be reducing the accuracy of the studies. For experiments investigating the effect of survival factors in chronic lymphocytic leukaemia it was considered important to be studying a purer population of what will now be referred to as B-CLL cells. Accordingly, a T cell depletion step was incorporated into the lymphocyte isolation procedure. Further work in this chapter describes an investigation into the finding that B-CLL cells have large amounts of caspase-8 which is not cleaved to any great extent during induction of apoptosis (see Chapter 4, section 4.5). Since caspase-8 is known to play a major role in the Fas signalling pathway, the response of B-CLL cells to stimulation of this pathway was investigated. 5.2 Purified CLL B lymphocytes are more sensitive to apoptosis in in vitro culture than unpurified populations of total CLL lymphocytes To further increase specificity when analysing apoptosis sensitivity of CLL cells, it was decided to purify the B lymphocyte fraction (B-CLL) from the total lymphocytes obtained from the patients. By incubating a proportion of the isolated lymphocytes with CD3+ dynabeads (Dynal, Oslo, Norway) according to the manufacturer s instructions, depletion of T cells was performed. A determination of the purity of the pre- and post-depletion fractions was obtained flow cytometrically using antibodies to CD3 and CD 19 (see Chapter 2, section 2.3.2) (Figure 5.1 A). 69

80 Chapter 5 Results 3 Blood samples were obtained from six patients and the lymphocyte fraction was isolated as described in chapter 2. A proportion of these cells were cultured for 24 h alone or in the presence of chlorambucil (7 pm) or staurosporine (0.2 pm). The remaining cells from each isolation were further purified by T cell depletion prior to being cultured for 24 h alone or in the presence of chlorambucil (7 pm) or staurosporine (0.2 pm). After the 24 h culture period, cells from all of the cultures were labelled using the Annexin V technique and analysed flow cytometrically. In all six cases, the level of spontaneous apoptosis in the purified B cell cultures was slightly increased over that in the total lymphocyte cultures (Figure 5.1 B), which does indicate that the B-CLL cells may be losing one or more survival stimuli as a result of the T cell depletion step. 70

81 Chapter 5 Results 3 Total lymphocyte population Post-T cell depletion o FL1 -Height (3) v s FL2-Height (4) 11.6% -T cell 1.3% -T cell - 8. W Ph I cn Q CJ. 83.4% B cells *92% B cells " l CD19-FITC 10 W io2 10J B 70 W B 60 2 C L o 50 Q. < w 40 o \ J m B cells Total cells 15 A 21A 23A 24A 25A Patient number 28A Figure 5.1 (A) Use of CD3+ magnetic beads enriches for CD 19+ B cells in cultures of CLL lymphocytes. (B) Purified B cells are more sensitive to spontaneous apoptosis than cultures of mixed lymphocytes. Lymphocytes were isolated from CLL blood samples and were split into duplicate cultures. T cells were depleted from one fraction using CD3+ magnetic beads. Both fractions of cells were cultured for 24 h prior to being labelled using the Annexin V technique to analyse the level of apoptosis in the samples. 71

82 Chapter 5 Results Culture of B-CLL cells with Interleukin-4 and CD40 stimulation results in a reduction in spontaneous apoptosis In order to investigate the requirement of B-CLL cells for stimulation by survival factors, two B cell growth factors, interleukin-4 and CD40, were incorporated into the study. Both of these factors had previously been identified as B cell growth stimulants in normal and malignant cells (Nakanishi et al, 1996; Crawford et al, 1993). Interleukin-4 has been reported to have anti-apoptotic effects when applied to in vitro cultures of B cells (Panayiotidis et al, 1993) and acts as an apoptosis inhibitory agent by maintaining the Bcl-2 expression level in B-CLL cells (Danescu et al, 1992). CD40 stimulation promotes survival by activation of NFkB (Rothe et al, 1995) and protein tyrosine phosphorylation (Laytragoon-Lewin et al, 1998). Freshly isolated, purified B-CLL cells from 10 patients were cultured alone or in the presence of recombinant human interleukin-4 (10 ng/ml). In all instances, the addition of interleukin-4 reduced the level of spontaneous apoptosis after 24 h of culture (figure 5.2 A), and increased viability of IL-4-treated cells was maintained to 48 h (figure 5.2 B). The effect of interleukin-4 on inhibition of spontaneous apoptosis varied between patient samples, but a reduction in spontaneous apoptosis level was seen in all of the samples. Addition of a monoclonal antibody directed against CD40 also reduced the level of spontaneous apoptosis in the majority of cases, however, in the cells from patient 27A, addition of anti-cd40 monoclonal antibody actually increased the level of spontaneous apoptosis (figure 5.3). Overall, the level of protection afforded by CD40 stimulation against spontaneous apoptosis was less dramatic than the effect provided by interleukin-4 stimulation. In order to investigate the combined effect of these two survival stimuli on B-CLL cell survival, samples of cells from eight patients were cultured with a combination of interleukin-4 (10 ng/ml) and anti-cd40 monoclonal antibody (1.5 pg/ml). The level of apoptosis in the cultures was assessed after 24 h using the Annexin V assay. In cases 23A, 27A, 29A and 31 A, anti-cd40 increased the protection provided by interleukin-4, in the remaining 4 cases, anti-cd40 abrogated the protective effect of interleukin-4 (figure 5.3). This trend may indicate a variable dependency on survival factors between patients. 72

83 Chapter 5 Results 3 A a Control B 10ng/ml IL4 15A 23A 27A 28A 29A 30A 31A 32A Patient Number B 100 = 60 B.2 > /in Control IL Oh 24 h 48 h Time (h) Figure 5.2 Interleukin-4 inhibits the induction of spontaneous apoptosis in purified B- CLL cell cultures. (A) Freshly purified cells were cultured for 24 h alone (control) or in the presence of 10 ng/ml IL-4. Apoptosis was assessed using the Annexin V assay to measure externalised phosphatidylserine. (B) B-CLL cells were cultured alone or in the presence of IL-4 (10 ng/ml) for 48 h. Samples (1 x 106 cells) were taken from each culture at 24 and 48 h following culture setup, viability was assessed by propidium iodide exclusion. Results plotted are the mean of six experiments. Standard error bars are also depicted. 73

84 Chapter 5 Results _co 50 Control IL4 H anti-cd40 m anti-cd40 + IL4 23A 24A 27A 28A 29A 30A 31A 32A Patient Number Figure 5.3 The effect of stimulation of B-CLL cells with interleukin-4 and antibodies to CD40 is case dependent. B-CLL cells were cultured for 24 h alone or in the presence of IL-4 (10 ng/ml), anti-cd40 monoclonal antibody (1.5 pg/ml) or both stimuli. Following the culture period, 1 x 106 cells from each culture were analysed for apoptosis level using the Annexin V assay. 74

85 Chapter 5 Results Culture of B-CLL cells with Interleukin-4 or CD40 stimulation increases their resistance to chemotherapeutic drugs In order to determine the level of protection afforded by interleukin-4 against apoptosis induced by chemotherapeutic drugs, B-CLL cells from patients 28A and 29 A were pre-incubated for 24 h alone (control) or in the presence of interleukin-4 (10 ng/ml). After 24 h pre-incubation, chlorambucil (7 pm) or staurosporine (0.2 pm) were added and the cells were cultured for a further 24 h. The protective effect of interleukin-4 against drug-induced apoptosis was then analysed by assessing the extent of apoptosis in these cultures compared to apoptosis levels in cultures of cells incubated for 24 h with chlorambucil or staurosporine alone. In both cases the addition of interleukin-4 to the cells reduced the level of spontaneous apoptosis. The level of apoptosis induced by chlorambucil and staurosporine was also reduced in the cells pre-incubated with interleukin-4, compared to the levels in cultures after 24 h incubation with the drugs alone. In cells from patient 28 A, the protection from druginduced apoptosis was no more than 7%. However, in cells from patient 29A, the level of apoptosis in the cultures pre-incubated with interleukin-4 were reduced by as much as 22% (figure 5.4). This finding, although concerned with a limited sample of B-CLL patients, again indicates that B-CLL cells from different patients vary in their level of dependence on survival factors, and also demonstrates the important role that interleukin-4 may play in maintaining survival of the malignant lymphocyte clone and protecting the cells from chemotherapeutic drug-induced apoptosis in vivo. In order to determine the protection from drug-induced apoptosis afforded by a combination of CD40 and interleukin-4 stimulation, B-CLL cells from three patients were cultured for 24 h in the presence of anti-cd40 mab (1.5 pg/ml) and IL-4 (10 ng/ml). Following this pre-incubation period, chlorambucil (7 pm) was added to the cultures. The cells were incubated for a further 24 h, prior to assessment of apoptosis levels in the cultures by Annexin V labelling. Control cultures were also analysed, where the cells had been incubated for the full 48 h in the presence of IL-4 and anti- CD40, without the addition of chlorambucil. 75

86 Chapter 5 Results 3 Spontaneous apoptosis was inhibited in all the three patient samples analysed, and chlorambucil-induced apoptosis was also inhibited in the samples which had been preincubated with IL-4 and anti-cd40 (Figure 5.5 A). The average protective effect of IL-4 + anti-cd40 against spontaneous apoptosis was 21.2%, and the average protection against chlorambucil-induced apoptosis was 26.8% (Figure 5.5 B). However, there was wide variation between the patient samples as to the effectiveness of IL-4 and CD40 stimulation. Cells from patients 21A and 27A were highly receptive to the protective effects of IL-4 and CD40 against chlorambucil-induced apoptosis, whereas the cells from patient 24A were not protected to such a great extent (Figure 5.5 A). The extent of protection against chlorambucil-induced apoptosis afforded by the two stimuli was not always to the same degree as the protective effect against spontaneous apoptosis. Cells from patient 21A were protected to a similar degree from spontaneous and chlorambucil-induced apoptosis, cells from patient 24A were protected to a greater degree from spontaneous apoptosis than from chlorambucilinduced apoptosis, and, interestingly, cells from patient 27A were protected to a greater extent against chlorambucil-induced apoptosis than against spontaneous apoptosis (figure 5.5 A). This again underlines the wide variability in survival factor dependency between individual B-CLL cases, a dependency which may also have implications in determining the response of different patients to chemotherapy. 76

87 Chapter 5 Results 3 A Patient No. Control IL-4 + CD40 Chi IL-4 + CD40 + Chi 21 A A A B 90 Control CD40+IL4 CHL CD40+IL4+CHL Figure 5.5 Stimulation of B-CLL cells with anti-cd40 and interleukin-4 protects against apoptosis induced by chlorambucil. B-CLL cells from three patients were cultured for 24 h alone (Control), or in the presence of chlorambucil (7 pm) (CHL), prior to assessment of apoptosis using the Annexin V assay. Identical cultures of cells from the same patients were pre-incubated for 24 h in the presence of anti-cd40 monoclonal antibody (1.5 pg/ml) and interleukin-4 (10 ng/ml) (CD40 + IL-4) prior to chlorambucil (7 pm) being added (CD40 + IL-4 + CHL) and the cells being cultured for a further 24 h. A. Table of results obtained from each patient analysed in this way. All numerical values are percentage apoptosis. B. Plotted mean results from the three patients. Standard error bars are also depicted. 77

88 Chapter 5 Results B-CLL cells are not sensitive to Fas-induced apoptosis High levels of the initator caspase, caspase-8 had previously been observed in B-CLL cells (see chapter 4, section 4.5). Caspase-8 cleavage in response to induction of apoptosis by chlorambucil was also observed (figure 4.5 A, lanes 7-9 and 4.5 B, lane 4). The role of caspase-8 activation in drug-induced apoptosis is unclear, but in proportion to the amount of caspase-8 pro-form present in these cells, there appeared to be little activation taking place as a result of treating the cells with chemotherapeutic drugs (figure 4.5). Since caspase-8 is known to play a major role in the Fas signalling pathway, the response of B-CLL cells to stimulation of this pathway using an anti-fas (IgM) monoclonal antibody was investigated. B-CLL cells isolated from eight patients were cultured alone or in the presence of anti-fas monoclonal antibody (clone CH11, 0.5 pg/ml) for 24 h. Following the incubation period, 1 x 106 cells from each culture were labelled using the Annexin V assay, and analysed flow cytometrically (figure 5.6). Fas stimulation induced apoptosis above the level of spontaneous apoptosis in only two instances (patients 22A and 29A). In the other six cases addition of anti-fas monoclonal antibody inhibited the induction of spontaneous apoptosis. This finding mirrors that of Laytragoon-Lewin and co-workers (1998) who also determined that B-CLL cells were generally not sensitive to Fas stimulation. Further studies were initiated in order to determine the nature of Fas resistance in B-CLL cells. 78

89 Chapter 5 Results 3 </) 8 Q. O * Control 24 h Fas 24 h 22A 23A 24A 25A 26A 27A 28A 29A Patient Number Figure 5.6 B-CLL cells are resistant to apoptosis induced by anti-fas monoclonal antibody. B-CLL cells were cultured for 24 h alone or in the presence of anti-fas mab (0.5 pg/ml). Analysis of apoptosis levels in the cultures was made using the Annexin V assay. 79

90 Chapter 5 Results Upregulation of Fas receptor on B-CLL cells does not increase sensitivity to apoptosis induced by Fas ligation. Previous reports have confirmed the low density of expression of the Fas receptor on B-CLL cells (Mainou-Fowler et al, 1995). Freshly isolated B-CLL cells were labelled with anti-fas monoclonal antibody and a FITC-conjugated secondary antibody, as described in Chapter 2, and analysed flow cytometrically for the level of expression of Fas receptor. Previous reports had demonstrated that Fas receptor levels on B-CLL cells could be upregulated by culturing the cells with either CD40 or interleukin-2 and Staphylococcus aureus Cowan I (Wang et al, 1997). To induce upregulated expression of Fas receptor, B-CLL cells were cultured for 24 h in the presence of anti-cd40 monoclonal antibody (1.5 pg/ml) alone or in combination with interleukin-4 (10 ng/ml). Since culturing B-CLL cells for 24 h (with or without CD40 stimulation) results in induction of apoptosis (see figure 5.3), interleukin-4 was added to the cultures in order to inhibit apoptosis to a level from which analysis of Fas-induced apoptosis could be made following on from the initial 24 h culture to upregulate Fas receptor expression. The level of Fas receptor on the freshly isolated cells was compared with the level of Fas receptor on the CD40 +/- IL-4 -stimulated CLL cells (figure 5.7 A). For all five patient samples, the level of expression of Fas receptor increased following stimulation of the cells with CD40, compared to the level of Fas receptor on the freshly isolated cells (figure 5.7 B). The addition of interleukin-4 did not appear to alter the effectiveness of CD40-induced Fas receptor upregulation (figure 5.7 B). 80

91 Chapter 5 Results 3 A. Control IgM (MFI = 2.3) / Freshly isolated B-CLL cells (MFI = 8.2) B-CLL cells + anti-cd40 mab (24 h) ^(M FI = 27.2) Log green fluorescence (FITC) B. Fresh CD40 24h CD40 + IL4 24h 21A 23A 24A 27A 29A 30A 31A 32A Patient Number Figure 5.7 Fas receptor expression is elevated on B-CLL cells following stimulation with antibodies to CD40 and/or interleukin-4. A. B-CLL cells (from patient 30A) show increased Fas receptor expression following 24h incubation with anti-cd40 mab (0.5 fig/ml) when compared with the level of expression on the freshly isolated cells. B. Culture of B-CLL cells with anti-cd40 alone or in combination with interleukin-4 results in upregulation of Fas receptor expression. Freshly isolated B-CLL cells were labelled with anti-fas monoclonal antibody (as described in Chapter 2, section 2.11) and analysed flow cytometrically for the amount of expression of the cell surface receptor, Fas. Subsequently, cells from the same patients were cultured for 24 h in the presence of anti-cd40 monoclonal antibody alone or with interleukin-4 (10 ng/ml) added, before analysis of Fas receptor expression was made again. 81

92 Chapter 5 Results 3 To determine if the increased expression of Fas receptor on the B-CLL cells would confer increased sensitivity to Fas-induced apoptosis, B-CLL cells from three patients which had been cultured with anti-cd40 (1.5 pg/ml) and IL-4 (10 ng/ml) for a period of 24 h to enable upregulation of Fas receptor expression as described above, were exposed to an anti-fas monoclonal antibody (clone CH-11, 0.5 pg/ml). Following a further 24 h culture period, the level of apoptosis in the cultures was analysed using the Annexin V assay. From the three cases analysed using this method, the level of apoptosis induced by anti-fas was not increased over that in the control (anti-cd40 + IL-4 alone) cultures (figure 5.8). To check that the addition of interleukin-4 to the cultures was not inducing Fas resistance, cells from patient 27A were cultured for 24 h with anti-cd40 alone prior to addition of anti-fas and a further 24 h culture period. At 48 h, the level of apoptosis in the CD40 alone culture was 59.7%, and the level of apoptosis in the culture pre-incubated with anti-cd40 and with anti-fas added for the second 24 h, was 57.1%. This analysis demonstrated that CD40 stimulation alone, whilst effectively upregulating Fas receptor, was not conferring Fas sensitivity on the B-CLL cells. It also demonstrated that the addition of interleukin-4 to the cultures for the first 24 h was not contributing to this resistance to Fas-induced apoptosis. 82

93 Chapter 5 Results ao Patient 21 A Time (h) 70 T 60.. Patient 29 A Time (h) 60 Patient 27 A Control a g Time (h) A CD95 CD40 + IL4 CD40+ IL4 + Fas Figure 5.8 Upregulation of Fas receptor on B-CLL cells does not increase their sensitivity to apoptosis induced by Fas stimulation. B-CLL cells from three patients were cultured for 24 h in the presence of anti-cd40 monoclonal antibody (1.5 fig/ml) and IL-4 (10 ng/ml). Subsequently, anti-fas antibody was added (CH11, 0.5 pg/ml). The cells were cultured for a further 24 h, before the extent of phosphatidylserine extemalisation in the cultures was assessed using the Annexin V assay. Results were compared against samples of cells stimulated for the 48 h period with anti-cd40 + IL- 4 alone, control cells which had received no stimulation, and cells which had received stimulation from anti-fas antibody for 48 h. 83

94 Chapter 5 Results 3 80 n 70 - co O 40 - Q _ < 30 - Control Fas tic CD40 m - CD40 + Fas Time (h) Figure 5.9 Stimulation of B-CLL cells with CD40 does not confer sensitivity to Fasinduced apoptosis. Cells from patient 27A were cultured for 24 h in the presence of anti-cd40 monoclonal antibody (1.5 pg/ml). Fas receptor upregulation was checked flow cytometrically (see figure 5.7B), anti-fas monoclonal antibody (clone CHI 1, 0.5 pg/ml) was added and the cells were cultured for a further 24 h (CD40 + Fas). Samples of cells from the same patient were cultured for 48 h alone (Control), with Fas stimulation alone (Fas) or with CD40 stimulation alone (CD40). The apoptosis level at 48 h in each of the cultures was assessed using the Annexin V assay to measure externalised phosphatidylserine. 84

95 Chapter 5 Results B-CLL cells do not overexpress the caspase-8 inhibitory protein, c-flip Several proteins have been identified which can inhibit Fas-induced apoptosis by blocking the interaction of caspase-8 with the death effector domain (DED) of the adapter molecule FADD. One such protein is c-flip. c-flip is present in mammalian cells as two isoforms, c-flip l (-55 kd) and c-flips ( 33 kd) (Irmler et al, 1997; Rasper et al, 1998). High levels of c-flip have been detected in some melanoma tumours, and c-flip is predominantly expressed in lymphoid and muscle tissue (Irmler et al, 1997), indicating that this protein may play a role in regulating apoptosis of lymphoid malignancies. The following experiments were performed in order to determine the relative expression of c-flip in B-CLL cells, compared to the expression levels of caspase-8, and to examine any alterations in expression as the cells were exposed to chemotherapeutic drugs and Fas stimulation. B-CLL cells from patients 21 A, 31A and 32A which had been cultured alone or in the presence of chlorambucil (7 pm), interleukin-4 (10 ng/ml) or anti-cd40 (1.5 pg/ml) were labelled using the Annexin V method, and analysed flow cytometrically to record the level of apoptosis in the cultures (figure 5.9 A). Samples of cells from these cultures were examined using immunoblotting for the presence of the caspase-8 inhibitory protein c-flip (figure 5.9 B). Blots were subsequently stripped, and re-probed with anti-caspase-8 polyclonal antibody (figure 5.9 C), allowing a comparison of the relative levels of c-flip and caspase-8 to be made. The immunoblot results demonstrate that B-CLL cells do express c-flip and caspase-8, but the level of expression of c-flip does not appear to be increased over that of caspase-8. This implies that c-flip overexpression is not responsible for the apoptotic block in Fas-induced apoptosis in B-CLL cells. Upon examination of the immunoblot results, it was noted that in one case (21 A) chlorambucil had induced clipping of the short form of c-flip (figure 5.9 B, lane 3), which occured concurrently with activation of caspase-8 (figure 5.9 C, lane 3). The positive control lane for this experiment was THP-1 monocytic cells treated for 4 h with anti-fas monoclonal antibody. In the positive control lane, c-flip is also cleaved. This phenomenon may represent a caspase dependent cleavage of c-flip. 85

96 A Patient 21A Patient 31A Patient 32A t 55 k D -^ '«* 4 3 k D ->» 28kD -> * 18kD->- Figure 5.7 B-CLL cells do not overexpress the caspase-8 inhibitory protein c-flip. Freshly isolated B-CLL cells (1) from patients 21 A, 31A and 32 A were cultured in vitro alone (2) or in the presence of chlorambucil (7 J»M) (3), interleukin-4 (10 ng/ml) (4) or anti-cd40 monoclonal antibody (1.5 jpg/ml) (5). (A) The level of apoptosis in the cultures was assessed after 24 h using the Annexin V assay. (B) Cells from the cultures were analysed by immunoblotting for the presence of c-flip. (C) The blots were stripped and re-probed with an antibody against caspase-8.

97 Chapter 5 Results CLL cells have all the necessary components to form a death inducing signalling complex (DISC) upon Fas ligation but do not assemble a DISC upon Fas stimulation The formation of a death-inducing signalling complex (DISC) in response to Fas ligation on Jurkat cells was reported in 1997 (Medema et al). The DISC consists of the intracellular death domains (DD) of the trimerised Fas receptors, the adapter molecule FADD, which binds to the Fas receptor DD through its own death domain. FADD also consists of a region called a death effector domain (DED) which recruits caspase-8 to the DISC. Cells which respond to Fas stimulation by assembling a DISC can be split into two categories. Type I cells assemble the DISC in a matter of seconds, whilst type II cells take minutes to form a DISC and activate downstream caspases such as caspase-3 (Scaffidi et al, 1998). In order to investigate the possible reasons for the resistance of B-CLL cells to Fas-mediated apoptosis, immunoblotting for FADD was performed on freshly isolated and cultured B-CLL cells (figure 5.10), the presence of caspase-8 already having been confirmed in these cells (figures 4.5 and 5.9). The expression level of FADD appeared to increase slightly after the culture period of 24 h, but no significant alterations in expression of FADD were seen as a result of culturing the cells with chlorambucil, anti-fas or anti- CD40 monoclonal antibodies or interleukin-4 (figure 5.10). In the cells from patient 22 A, there is evidence of a slight decrease in FADD expression following treatment of the cells with staurosporine (figure 5.10 A, lane 4), which corresponded with induction of a high level of apoptosis and activation of caspase-8 (figure 5.10 B, lane 4). In samples from patient 22 A, FADD appeared as a single band at approximately 26 kd. In all other patient samples, FADD appeared as a doublet band (Figure 5.10 C and D), which may correspond to different FADD isoforms, or serine phosphorylation states (Xerri et al, 1999). SKW6.4 (murine B cells, Type I) and Jurkat (human T cells, Type II) cell lines were used as positive controls for DISC formation and isolation. Using the method of Peter and co-workers (1998) as described in chapter 2, section 2.12, SKW6.4 and Jurkat cells were stimulated with anti-fas monoclonal antibody (anti-apo-1, IgG) for 15 minutes. 87

98 Chapter 5 Results 3 24 h B?4h f ^ Oh Con Chi STS Fas Oh C o n C hi STS Fas 55/53 * * mm mm F A D D > - A k D > mm ~ 26 kd 28 kd > 18 kd > % A po 24 h Thp-1 O h rcon C hi IL 4 C D 40 + STS D Patient Number Single or double band corresponding to FADD 15 A Double 21 A Double FADD > 9 ~ 26 kd 22 A 28 A 29 A Single Double Double 31 A 32 A Double Double Figure 5.10 B-CLL cells express the adapter protein, FADD. (A) Freshly isolated B-CLL cells from patient 22 A, and those cultured for 24 h alone (Con) or with chlorambucil (Chi, 7 pm), staurosporine (STS, 0.2 pm), or anti-fas monoclonal antibody (Fas, 0.5 pg/ml) as indicated, were run on a 4-12% gradient polyacrylamide gel, blotted onto nitrocellulose filter and probed with a monoclonal antibody directed against the DISC adapter molecule, FADD. In patient 22 A, FADD appears as a single band at approximately 26 kd. (B) Samples of cells from the same patient were run on a gel, blotted and probed with a polyclonal antibody to caspase-8, as described previously. The percentage apoptosis level in the cultures is indicated below the blot. (C) Cells from patient 32 A, were cultured alone (Con), or with chlorambucil (Chi, 7 pm), interleukin-4 (IL4, 10 ng/ml) or anti-cd40 monoclonal antibody (CD40, 1.5 pg/ml), prior to being run on a 12 % polyacrylamide gel, blotted onto nitrocellulose and probed with a monoclonal antibody directed against FADD. In this patient, FADD appears as a doublet band. (D) Table showing FADD expression in all 7 cases analysed. FADD appears as a doublet band in the majority of cases. 88

99 Chapter 5 Results 3 Samples of stimulated and unstimulated (control) cells were lysed, and the same concentration of anti-fas antibody used for stimulation was added to the control samples as a negative control. The cells were incubated with Protein A-Sepharose to immunoprecipitate any complexes which may have formed as a result of stimulating the Fas receptor with an IgG anti-apo-1 monoclonal antibody. Immunoblotting of immunoprecipitates of anti-fas stimulated cells showed a band corresponding to FADD (~26 kd), in the lane containing stimulated SKW6.4 cells, which was absent in the lane containing stimulated Jurkat cells (figure 5.12A, lanes 1 & 3). This demonstrated that FADD had been recruited to a DISC within 15 minutes in SKW6.4 cells, but not in Jurkat cells. The negative control lanes (lysate supplemented samples, figure 5.12A, lanes 2 & 4) showed higher levels of FADD than was evident in the lanes containing stimulated cells. This was most likely due to inefficient cell lysis leaving functional, membrane bound Fas receptor in the lysate which could be stimulated to form DISCs. FADD molecules which are also freely available in the lysate appear to bind more effectively in this situation. In order to investigate whether or not Fas-stimulated B-CLL cells were capable of forming a DISC, similar experiments on B-CLL cells were performed which included SKW6.4 cells as a positive control. B-CLL cells from patients 31A and 32A were cultured for 24 h in the presence of anti-cd40 in order to upregulate Fas receptor expression, which was checked flow cytometrically. The cells were subsequently treated with anti-fas monoclonal antibody (anti-apo-1, IgG) for 60 minutes. Following lysis of the cells the anti-fas bound complex was immunoprecipitated using Protein A-Sepharose, subsequently the DISC complex was immunoblotted for FADD and caspase-8. SKW6.4 cells assembled FADD to the DISC after only 15 minutes (figure 5.12A), however, after stimulation with anti-fas monoclonal antibody for 60 minutes, B-CLL cells did not show FADD binding to a DISC (figure 5.12B). The blots were stripped and re-probed with anti-caspase-8 polyclonal antibody, however, due to the presence of a strong band on the blots corresponding to protein A (~42 kd), the presence of caspase-8 proform and activated fragments (55 and 43 kd) could not be established (figure 5.13). 89

100 Chapter 5 Results 3 c l CL 3 C/3 S 3 B ^ C/3 \ g +T* I X rr> C /3 & i4 C /D c3 1 3 o. D. 3co S C3 c3 i 3 IgG FADD (26 kd) B C3 -j S ' - P c 3 a. a. 3 C/3 & C3 CO CO V >4 * C /3 C /3 C /3 J V CQ I 3 J u CQ E.g VP to 3 -<1-3 M5 * C /1 c / 3 S 1 3 t o hj _l u CQ o CQ < IgG I g G > < - FADD (26 kd) - > «*» * * Patient 32 A Patient 31 A Figure Immunoblotting for FADD on immunoprecipitates of Fas-stimulated cells. (A) SKW6.4 and Jurkat cells were stimulated for 15 minutes with anti-fas monoclonal antibody (anti-apo-1, IgG). Samples of stimulated and unstimulated cells were lysed, and anti-apo-1 was added to the unstimulated lysates as a control. The protein complexes formed by Fas stimulation were immunoprecipitated using Protein-A-Sepharose. The lysates were run on a 4-12% gradient SDS-polyacrylamide gel, and transferred onto a nitrocellulose filter. The filter was probed with a monoclonal antibody directed against FADD. (B). B-CLL cells from patients 31A and 32A were stimulated for 60 minutes with anti- Fas (anti-apo-1, IgG). Lysates from stimulated and unstimulated cells, along with samples of SKW6.4 cells treated as above, were run on 4-12% polyacrylamide gels, transferred onto nitrocellulose filters and probed with anti-fadd monoclonal antibody. 90

101 Chapter 5 Results 3 ts * 3 i > JU o a. a. 3 C/5 E g M d on ' & c3 C/5 C C/5 a on 3 3 VO J _ ) J u i t V V C/5 C/5 C/5 0Q 0Q 55 kd ^ " 43 k D ->» * FADD (26 k D )> k D > Figure 5.13 Immunoblotting for caspase-8 on lysates of anti-fas stimulated cells. Cells from patient 32A were stimulated with anti-apo-1 for 60 minutes, lysed and incubated with Protein-A-Sepharose for 90 minutes. Samples of stimulated and unstimulated B-CLL cells were run on a 4-12 % polyacrylamide gel along with samples of stimulated (15 minutes), unstimulated and lysate supplemented samples of SKW6.4 cells. The proteins were transferred onto a nitrocellulose filter, which was probed with a polyclonal antibody to caspase-8 (as described previously). 91

102 Chapter 5 Results Discussion Survival factors in B-CLL In order to increase specificity when analysing apoptosis in CLL, it was decided to purify B-CLL cells from the total lymphocyte fraction, as had been used in previous experiments. B-CLL patients have a high proportion of tumour B cells in their peripheral blood compared with the ratio of T to B cells in normal blood. The patients sampled in this study had white blood cell counts ranging from 3.7 to 68 x 109/L, the average white cell count being 28 x 109/L. In cases such as these the T cell fraction can be assumed to be minimal, but nevertheless still present. For studies into the effect of growth factors of B-CLL cells, the presence of T cells may provide a contaminating source of survival stimuli and so removal of this fraction was deemed necessary. Accordingly, a T cell depletion step was incorporated into the B-CLL cell purification procedure. In order to assess the degree to which T cell-mediated stimuli could influence B-CLL cell survival, a series of comparitive experiments were performed. In all six of the cases analysed, the level of spontaneous apoptosis was elevated in the purer B-CLL culture compared to the culture containing B and T cells (Figure 5.1 B). The difference in sensitivity was significant in only one case, patient 24A, where the B-CLL cells were 21.9% more sensitive to spontaneous apoptosis than cells in the unpurified culture. In the remaining cases the difference was slight (4.6% average), but still demonstrates the effect that T cells may have in stimulating survival of B-CLL cells. The effects on cell growth and apoptosis of growth factor additions to cultures of CLL cells has been investigated by a number of groups, (DeFrance et al, 1991, Mainou- Fowler et al, 1995) but information regarding variable requirements between cases is limited. In order to investigate the requirement of B-CLL cells for stimulation by survival factors, the effects of two B cell growth factors, interleukin-4 and CD40, were studied. Both of these factors had previously been identified as B cell growth stimulants in normal and malignant cells (Nakanishi et al, 1996; Crawford et al, 1993). 92

103 Chapter 5 Results 3 Interleukin-4 had previously been reported to have anti-apoptotic effects when applied to in vitro cultures of B cells (Panayiotidis et al, 1993) and so the ability of interleukin-4 to inhibit spontaneous apoptosis of B-CLL cells was investigated. Interleukin-4 inhibited spontaneous apoptosis in all of the patient samples analysed, and could significantly improve the survival of the cells (Figure 5.2A and B). The extent of inhibition of spontaneous apoptosis ranged from 5% to 53.8%, indicating that B-CLL cells vary in their dependency on interleukin-4 for protection against apoptosis induction. In order to investigate this further, it would have been interesting to monitor the Bcl-2 expression levels in these cells before and after culture with interleukin-4, since this is one of the ways in which IL-4 might protect cells from apoptosis (Danescu et al, 1992). An inter-patient variability in sensitivity to interleukin-4 stimulation has been described above. It was of interest, therefore, to discover if IL-4 sensitive B-CLL cells could be equally as sensitive to a second survival stimulus, and to determine whether an additional survival advantage could be provided by stimulation of the cells with a combination of growth factors. A second B cell survival factor is CD40. CD40 stimulation promotes survival by activation of the anti-apoptotic transcription factor, NFkP (Rothe et al, 1995) and protein tyrosine phosphorylation (Laytragoon-Lewin et al, 1998). In this study, samples of B-CLL cells from eight patients were cultured in the presence of interleukin-4 and a monoclonal antibody to CD40. This culture system for delivery of the CD40 stimulation was as described by Dive et al, This analysis revealed that B-CLL cells were less sensitive to CD40 stimulation than to interleukin-4 stimulation (Figure 5.3). In fact, in one of the patient samples analysed, addition of anti-cd40 monoclonal antibody induced apoptosis above the level in the control culture, a phenomenon which had been reported previously (Wang et al, 1997). Stimulation of B-CLL cells with a combination of interleukin-4 and CD40 stimulation revealed further inter-patient variability. In 50% of the samples, the combination of survival factors reduced the level of apoptosis below that resulting from stimulation of the cells with interleukin-4 alone. In the remaining 50% of samples, the combination of stimuli was less effective at inhibiting spontaneous apoptosis than interleukin-4 alone. 93

104 Chapter 5 Results 3 Since B-CLL cells appear to vary in their sensitivity to inhibition of spontaneous apoptosis by growth factors, it would be interesting to discover if these stimuli could also promote resistance of the cells to apoptosis induced by chemotherapeutic drugs. The level of apoptosis in drug-treated cultures pre-incubated with IL-4 was less than the level of apoptosis in the cultures incubated with the drugs alone (Figure 5.4). Samples of cells from two patients were analysed for the protective effect of interleukin-4 against drug-induced apoptosis, and the cells from the two patients differed in the degree of protection afforded by interleukin-4. A combination of interleukin-4 and CD40 stimulation also inhibited induction of apoptosis by chlorambucil (Figure 5.5). Again, the response differed between cells from the three patient samples analysed, the percentage reduction in apoptosis level in the survival factor pre-incubated samples, compared to the cultures incubated with the drugs alone, ranging from 4.9% to 41.5%. Because this reduction in apoptotic rate might be due to inhibition of spontaneous apoptosis, and not due to inhibition of drug-induced apoptosis, samples from the same patients were cultured alone and with the survival stimuli alone as controls. Results from one patient (27A) showed that IL-4 and CD40 stimulation had inhibited drug-induced apoptosis to a much greater extent than spontaneous apoptosis, indicating that the survival factors were affecting the cells response to chemotherapeutic drugs. Of the remaining two cases, one showed the reverse pattern, indicating that in this patient s cells the survival stimuli were only effective in reducing spontaneous apoptosis, and in the third case, spontaneous and drug-induced apoptosis were inhibited to roughly the same extent. These findings demonstrate a wide inter-patient variability in dependence of the B- CLL cells on stimulation from survival stimuli. Spontaneous apoptosis and chemotherapeutic drug-induced apoptosis can be inhibited by culturing B-CLL cells with interleukin-4 and/or CD40 stimulation. This raises the question of how relevant these factors are in determining apoptosis sensitivity in vivo. B-CLL patients have already been shown to have elevated levels of CD40 ligand in their serum (Younes et al, 1998), and this, in conjunction with a normal level of expression of CD40 receptors on the CLL cells (Laytragoon-Lewin et al, 1998), could be one mechanism in which B-CLL cells escape apoptotic cell death. B-CLL cells have also been shown to express normal levels of the interleukin-4 receptor (Gileece et al, 1993), and so the 94

105 Chapter 5 Results 3 levels of interleukm-4 in the peripheral blood of B-CLL patients could also be of great importance. In relation to this, it has been shown that T cells derived from B-CLL patients have increased levels of cytoplasmic IL-4 compared to normal control T cells (Mu et al, 1997), although whether or not this IL-4 was secreted was not determined. Further studies could be performed, possibly using ELISA-based assays, which would determine whether or not this IL-4 is secreted by the B-CLL T cells, and whether or not B-CLL patients have normal or elevated levels of serum interleukin Investigations into the Fas signalling pathway in B-CLL cells Previous experiments in this study had demonstrated that B-CLL cells have caspase-8 proform present in abundance. However, caspase-8 does not seem to be cleaved to any great extent during drug-induced apoptosis (Figure 4.5). Caspase-8 is the apical caspase in the signalling pathway induced following stimulation of the Fas receptor. B-CLL cells are known to express reduced levels of Fas receptor, and this, in conjunction with high levels of inactive caspase-8 may be one mechanism whereby B- CLL cells are resistant to apoptosis. Investigations into the Fas signalling pathway in B-CLL cells began with a series of experiments to determine whether the cells could be triggered into apoptosis via Fas signalling. In only 2/8 patients did Fas stimulation induce apoptosis above the level of spontaneous apoptosis. In the remaining cases, Fas stimulation inhibited spontaneous apoptosis. One reason for the resistance of B-CLL cells to Fas-induced apoptosis may be the low level of expression of Fas receptor on the cells surface (Mainou-Fowler et al, 1995; Wang et al, 1997). There are a number of ways in which Fas receptor expression can be upregulated on B cells. Tonsillar B cells and Burkitt s Lymphoma cells can be induced to upregulate Fas receptor expression by CD40 stimulation (Garrone et al, 1995; Scattner et al, 1995), and B- CLL cells have been shown to upregulate Fas receptor upon stimulation with a combination of interleukin-2 and a mitogenic stimulus such as Staphylococcus Aureus Cowan I (Mapara et al, 1993). In all of these studies the cells became sensitive to Fas stimulation following upregulation of receptor levels. In this study, B-CLL cells were cultured for 24 h with a monoclonal antibody to CD40 in order to upregulate Fas receptor expression. Comparison of the level of Fas receptor on freshly isolated B- CLL cells with the levels on CD40 stimulated cells demonstrated that CD40 95

106 Chapter 5 Results 3 stimulation was indeed inducing increased expression of Fas receptor on these cells (Figure 5.6). The aim of upregulating Fas receptor levels was to analyse the functionality of the Fas signalling pathway in B-CLL cells. Following 24 h pre-incubation with CD40 stimulation, the cells were cultured for a further 24 h with anti-fas monoclonal antibody in order to assess Fas sensitivity. However, by the time the 48 h time point was reached, the level of spontaneous apoptosis in the control cultures was very high, as was the level of apoptosis in the CD40-stimulated samples. This meant that examination of the Fas signalling pathway would be hampered by a high baseline level of apoptosis. In order to promote survival of the B-CLL cells over the first 24 h culture period, interleukin-4 was added to the cultures. The combination of CD40 and interleukin-4 stimulation also resulted in upregulation of Fas receptor, but successfully inhibited the induction of high levels of spontaneous apoptosis making analysis of Fas-induced apoptosis possible. To determine whether or not the upregulated Fas receptor was functional, cells which had been cultured with CD40 and interleukin-4 for 24 h were stimulated with anti-fas monoclonal antibody for a further 24 h. After this time, apoptosis levels in the cultures were measured using the Annexin V assay. None of the three patient samples analysed in this manner were sensitised to Fas-induced apoptosis following CD40 plus interleukin-4 stimulation (Figure 5.8). Interleukin-4 had previously been shown to inhibit Fas-induced apoptosis in B-CLL cells (Mainou-Fowler et al, 1995). To determine whether the addition of interleukin-4 to the cultures for the first 24 h was causing resistance to Fas-induced apoptosis, cells from one patient were cultured with CD40 alone prior to fas stimulation. However, no increased sensitivity to Fas-induced apoptosis was seen (Figure 5.9). Reports that CD40-upregulated Fas receptor can be functional have been published. Stimulation of murine B cells with anti-cd40 results in upregulation of Fas receptor, and renders the cells sensitive to Fas-induced apoptosis (Nakanishi et al, 1996). When human tonsillar B cells are stimulated with anti-cd40 to upregulate Fas receptor, they too become sensitive to Fas-induced apoptosis (Garrone et al, 1995), as do Burkitt s 96

107 Chapter 5 Results 3 lymphoma cells (Scattner et al, 1995). However, the situation in B-CLL cells appears to differ somewhat. Wang and co-workers (1997) reported that, regardless of the stimulus used to upregulate Fas receptor on B-CLL cells (including CD40 stimulation or culture with interleukin-2 and pokeweed mitogen), the cells did not become sensitive to Fas-induced apoptosis. This result conflicts with that of Marpara and coworkers (1995) who demonstrated increased sensitivity to Fas-induced apoptosis in B- CLL cells pre-incubated with interleukin-2 and S. aureus Cowan I. The sample group of patients analysed in this study was very small, but this study has demonstrated that upregulation of Fas receptor by CD40 stimulation does not increase sensitivity to Fasinduced apoptosis in B-CLL cells. To further define the nature of Fas resistance in B-CLL cells, components of the signalling pathway upstream of caspase-8 activation were investigated. Immunoblotting for proteins involved in the formation of the death inducing signalling complex (DISC) was performed. This analysis confirmed that B-CLL cells contain the adapter molecule, FADD (Chinnaiyan et al, 1996), which is required for caspase-8 activation in response to stimulation of the Fas receptor (Boldin et al, 1996). Since B-CLL cells had now been shown to express both FADD and caspase-8, and since upregulation of the Fas receptor had been confirmed, it was decided to analyse expression of a known inhibitor of Fas signalling. c-flip (FLICE-like inhibitory protein) is an inhibitor of the Fas receptor/ligand system, and acts at the level of FADD/caspase-8 binding. Interaction of c-flip with activated caspase-8 causes cleavage and activation of c-flip, which results in the c-flip/caspase-8 interaction becoming more inhibitive. Since c-flip is predominantly expressed in lymphoid and muscle tissue (Irmler et al, 1997), this indicates that this protein may play a role in regulating apoptosis of lymphoid malignancies. To determine whether B-CLL cells were overexpressing c-flip, in comparison with the level of caspase-8, immunoblotting using a polyclonal antibody to c-flip (Rasper et al, 1998) was performed on B-CLL cells which had been stimulated with chlormabucil or anti-fas monoclonal antibody. The blots were re-probed with anti-caspase-8 polyclonal antibody, so that the relative expression of each protein could be compared. This analysis demonstrated that, while B-CLL cells do express c-flip, the level of expression does not appear to be elevated above that of caspase-8 (Figure 5.9), which 97

108 Chapter 5 Results 3 means that c-flip overexpression is unlikely to be responsible for the block in apoptosis signalling through Fas in B-CLL cells. Fas signalling occurs by two distinct mechanisms (Scaffidi et al, 1998). Type I Fas signalling is extremely rapid, occuring in under 10 minutes, and requires the formation of a death inducing signalling complex (DISC). Type II Fas signalling is slower, and requires amplification of the apoptotic signal through release of cytochrome c from the mitochondria, which activates the apoptosome, a complex consisting of caspase-9 and Apaf-1. Since B-CLL cells do not undergo apoptosis in response to Fas signalling, regardless of the level of expression of Fas receptor on the cells surface, it was postulated that the B-CLL cells may be unable to form a DISC to transmit the apoptotic stimulus into the cell. Accordingly, analysis of the ability of B- CLL cells to form DISC S was carried out. SKW 6.4 murine T cells were used as positive controls for Type I DISC formation, and Jurkat T cells were used as controls for the Type II Fas response. As expected, SKW 6.4 cells showed binding of FADD to a Fas-induced DISC after only 10 minutes of stimulation (Figure 5.13 A). B-CLL and Jurkat cells, however, did not show any evidence of DISC formation even after 60 minutes of Fas stimulation (Figure 5.13 B), confirming that B-CLL cells cannot be classified as Type I cells. Whilst this evidence does not fully reveal the nature of Fas resistance in B-CLL cells, it does throw up some interesting possibilities. The Type II signalling pathway, unlike the Type I pathway, can be inhibited by Bcl-2 (Scaffidi et al, 1998), most probably at the level of amplification of the apoptotic signal at the mitochondria leading to apoptosome activation. Several studies have correlated apoptotic resistance in B-CLL with increased expression of Bcl-2 (Thomas et al, 1996; Aguilar-Santelises et al, 1996; Pepper et al, 1996). This evidence in conjunction with the discovery that B-CLL cells are most likely to act in a Type II manner in response to Fas stimulation, may be one reason for apoptotic resistance in B-CLL cells. In addition, other groups have reported that some cytotoxic drugs, such as doxorubicin, can induce apoptosis via the Fas signalling pathway (Freisen et al, 1997; Fulda et al, 1997). If this pathway is inhibited in B-CLL by the overexpression of bcl- 2, this may account for some of the drug resistance evident in B-CLL patients. This present study was limited in the number of patient samples analysed, and further investigations could monitor expression levels of Bcl-2 in conjunction with analysis 98

109 Chapter 5 Results 3 into apoptosome formation, possibly by monitoring activation of caspase-9, to determine whether B-CLL cells respond to Fas signalling or stimulation with cytotoxic drugs in a Type II manner.

110 Chapter Six General Discussion Chapter Six - General Discussion and Suggestions for Future Work This thesis has described an investigation into the significance of apoptosis in B cell chronic lymphocytic leukaemia. The lymphoaccumulative nature of B-CLL implies that dysregulation of the apoptotic process may be responsible for the development and progression of the disease. Additionally, apoptosis is known to result from treatment of malignant cells with chemotherapeutic drugs. The fact that drug resistance is a major problem in CLL, indicates that there may be a problem in apoptosis induction in B-CLL cells. In order to investigate this relationship, preliminary work was performed in order to assess the application of techniques for analysing apoptosis to specimens of freshly isolated CLL cells. Use of flow cytometric techniques, supplemented with agarose gel electrophoretic methods allowed quantification of the percentage of apoptotic cells in any given sample of CLL cells, although the labelling techniques used altered as the study progressed and new techniques became available. Initial studies demonstrated a low in vivo level of apoptosis in B-CLL patients, and confirmed the existence of spontaneous apoptosis when freshly isolated B-CLL cells were cultured in vitro. Variations in the level of spontaneous apoptosis between cases indicated that B-CLL cells differed from patient to patient in their dependence on external survival stimuli, which were not provided by the standard in vitro culture environment. Preliminary findings in this study also showed that the sensitivity of B- CLL cells to drug-induced apoptosis was closely related to the sensitivity of the cells to spontaneous apoptosis. Taken together, these findings would appear to implicate the degree of survival factor dependency of the malignant cells as an important factor in determining the response of patients to chemotherapy, thus making this an important area for future research. Candidates for survival stimuli in B-CLL include anti-apoptotic members of the Bcl-2 family, type II cytokines (e.g. interleukin-4), and members of the TNF/NGF receptor/ligand superfamily, in particular CD40 and its ligand. The role of two of these factors (CD40 and interleukin-4) was investigated in this thesis, where it was shown that both factors could inhibit some degree of spontaneous or drug-induced apoptosis (see chapter 5). In relation to this, a 100

111 Chapter Six General Discussion combination of CD40 stimulation and overexpression of Bcl-2 has been shown to increase the clonogenic survival of chlorambucil-treated B-CLL cells, and as such may contribute towards the acquisition of drug resistance (Walker et al, 1997), a major problem in advanced B-CLL cases. In this study, the effects of only two B cell survival factors have been investigated. Other factors may be important in promoting survival of B cells. Enhanced survival of B-CLL cells when cultured in direct contact with bone marrow stromal cells has been reported, the effect mediated by pi and p2 integrins and linked to maintainence of Bcl-2 levels (Lagneaux et al, 1998). The authors of this report postulate that this contact derived survival stimulus could play a role in accumulation of B-CLL cells in the bone marrow. These findings and the results contained within this thesis demonstrate the extent to which B-CLL cells depend on extracellular stimuli for enhanced survival and escape from apoptosis. As discussed above, this may also be a crucial factor in determining the response of B-CLL cells to chemotherapy. Further investigations in this area should incorporate a wider selection of B cell survival factors and use a larger group of patient samples in order to better determine the significance of the findings. The relative levels of the factors in the peripheral blood and bone marrow of B-CLL patients, or the availability of such factors accessible through contact with other cell types could also be investigated. In addition, samples of B-CLL cells derived from bone marrow rather than peripheral blood may have an altered dependence on growth factor stimulation for survival. One way in which this may occur, the survival stimulus being derived from contact with bone marrow stromal cells, is mentioned above (Lagneaux et al, 1998). Analysis of bone marrow derived B-CLL cells could provide insights into the nature of more advanced or drug resistant B-CLL. Other investigations performed during this study focused on components of the apoptotic machinery. Proteases belonging to the caspase family are the machinery enzymes involved in the degradation of an apoptotic cell, and are related to the ced-3 death gene of the nematode worm, C. elegans. They reside in the cell as inactive zymogens which require cleavage at specific aspartate residues in order to attain their active form. Caspase-3 and caspase-7 are two of the main effector caspases, 101

112 Chapter Six General Discussion responsible for initiating the terminal events of the apoptotic cascade such as DNA degradation. Other caspases, including caspase-2, caspase-8 and caspase-9 have been termed initiator caspases due to their involvement in upstream events such as Fas/CD95 receptor mediated apoptosis and initiation of apoptosis via mitochondria. The important role that these proteases play is underlined by the finding that caspase- 3 null mice develop brain tumours. Also, it has recently been reported that in human lymphoma the cellular location of caspase-3 can be an important prognostic indicator (Donoghue et al, 1999). Work described in this thesis has demonstrated the expression of caspase-2, caspase-3, caspase-7 and caspase-8 in B-CLL cells, all of which, with the exception of caspase-2, are activated as a result of induction of apoptosis by the chemotherapeutic drugs, chlorambucil and prednisolone. It has been postulated that the inability of B-CLL cells to undergo apoptosis may have been due to the absence of one or more caspases, however, this would not appear to be the case. Caspase-8 pro-form was shown to be present in abundance, but was cleaved in only small amounts in response to chlorambucil treatment. Since this caspase is primarily involved in apoptosis mediated induced by the Fas/CD95 receptor, the ability of B- CLL cells to undergo apoptosis in response to Fas stimulation was subsequently investigated. This included an investigation into the expression of Fas inhibitory proteins, and an examination of the ability of Fas-stimulated B-CLL cells to form a death inducing signalling complex (DISC). In this study, a combination of CD40 and interleukin-4 was used to upregulate expression of the Fas receptor on B-CLL cells. Despite this fact, it appears that the choice of interleukin-4 may not have been prudent. A recent report has been published which describes a system for upregulating Fas receptor using interleukins -2, -7 and -12 in combination with lipopolysaccharide (LPS) (Williams et al, 1999). Fas receptor upregulated in this system was functional in that the Fas bearing cells could be lysed by effector cells expressing Fas ligand. The addition of interleukin-4 to the combination of stimulatory cytokines resulted in delayed upregulation of Fas receptor, which was not functional in that apoptosis could not be triggered by fas ligand. In the study presented in this thesis, although Fas receptor was upregulated by CD40 and IL-4 stimulation, no increase in sensitivity to Fas-induced apoptosis was evident, which may be due to the effects of interleukin-4. However, in addition to the effects of interleukin-4, there may be a second factor involved in determining Fas-mediated apoptosis sensitivity. The report described 102